A multi-energy complementary and multi-cycle nested optimization modeling and solution method for water, wind and light

Abstract

The present invention belongs to the technical field of modeling and solving for the simulation of complementary production of hydropower, wind power and photovoltaic power, and relates to a method for modeling and solving the multi-period nested optimization of complementary production of hydropower, wind power and photovoltaic power. First, on the premise of not destroying the time continuity of the hourly simulation results of the long period, according to the coupling relationship between the hydropower simulation results at different time resolutions, the original continuous multi-year hourly simulation is decomposed into annual, monthly and hourly simulation models. Secondly, based on the parallelogram two-dimensional interpolation method, without introducing 0-1 integer variables, a linear approximation model is established for the hydropower nonlinear function, and all the decomposed models are converted into linear programming models that can be quickly solved. The simulation runs in a rolling manner according to the time sequence, and a simulation result correction mechanism is added to feedback and update the unreasonable simulation results layer by layer to ensure the solvability of the large-scale long-time sequence operation simulation model. The present invention provides simulation support for more reasonably planning the new energy capacity for complementary operation with hydropower stations.

Description

Technical Field

[0001] The present invention belongs to the technical field of modeling and solving for the production simulation of multi-energy complementarity of water, wind and light, and relates to a method for modeling and solving multi-period nested optimization of multi-energy complementarity of water, wind and light. Background Art

[0002] To achieve the goal of power system decarbonization, renewable energy is developing rapidly globally. Wind power and photovoltaic power generation bring high variability and uncertainty to the power system, and greater power system flexibility is required to ensure strict real-time power supply and demand balance. Hydropower is a clean and economic technology that can compensate for the volatility of wind and light, and has advantages such as fast startup, fast ramping, and reservoir storage. The complementary operation of wind power, photovoltaic and existing hydropower stations has become an important way for the planning of large-scale new energy power stations.

[0003] Due to the variability of wind and light resources and runoff, as well as the complex non-linear operating characteristics of hydropower stations, how to determine the new energy capacity for complementary operation with hydropower stations is a real challenge. Wind and light resources and runoff have significant inter-annual variations, which makes the traditional operation simulation methods based on typical days or typical years unable to evaluate the long-term performance of the complementary operation of water, wind and light. However, continuous multi-year hourly simulations will greatly increase the difficulty of solving.

[0004] Firstly, the number of decision variables in the hydropower system is numerous, resulting in obvious large-scale and high-dimensional characteristics in hydropower operation simulation; secondly, most of the characteristic curves of hydropower stations (such as water level-storage capacity curves, NHQ curves, etc.) have non-linear characteristics, and traditional linearization means are difficult to handle a large number of complex non-convex functions; secondly, hydropower has multi-stage dynamic characteristics. For a single power station, the hydropower station strictly satisfies the time-series constraints of water balance. For cascade power stations, the time-series hydraulic connection between upstream and downstream power stations also needs to be considered. At present, methods such as dynamic programming, non-linear programming, and linear programming are widely used in the hydropower system. However, most of these methods need to calculate and store information such as all potential state combinations and their index values. Under the combined action of factors such as the number of power stations, the number of time period states, and the number of decisions, the computational amount and storage amount required by the algorithm will grow non-linearly with the system scale, greatly increasing the execution overhead of the algorithm and causing a very serious curse of dimensionality problem. Various artificial intelligence algorithms (such as genetic algorithms, ant colony algorithms, particle swarm algorithms, etc.) have the advantages of being simple and easy to implement, implicit parallelism, high applicability, and strong robustness, and are almost not troubled by the curse of dimensionality problem. However, the quality of the solutions of artificial intelligence algorithms is easily affected by the operation operators and related parameter settings, and the phenomenon of premature convergence is prominent, and the rationality, stability and feasibility of the results cannot be strictly guaranteed.

[0005] In summary, the operation of the hydropower system has characteristics such as high dimensionality, strong non-linearity, multiple calculation stages, tight spatio-temporal coupling, and complex scheduling requirements, resulting in a huge computational burden. The addition of new energy has further exacerbated the difficulty of solving mathematical optimization problems due to the complementary operation constraints of hydropower, wind, and solar. For the simulation of the complementary operation of hydropower, wind, and solar with a cycle of up to several years and a resolution accurate to the hourly scale, issues such as how to model, whether it can be solved, and how to solve it efficiently are urgent problems to be addressed. Summary of the Invention

[0006] The present invention combines a multi-resolution soft connection framework with a linearization modeling method to construct a multi-period nested optimization modeling and solving method for the complementary operation of hydropower, wind, and solar, providing simulation support for more reasonably determining the new energy capacity for complementary operation with hydropower stations. First, without disrupting the time continuity of the hourly simulation results of the long cycle, based on the coupling relationship between the hydropower simulation results at different time resolutions, the original continuous multi-year hourly simulation is decomposed into a series of smaller-scale annual, monthly, and hourly simulation models. Secondly, based on the parallelogram two-dimensional interpolation method, without introducing 0-1 integer variables, a linear approximation model is established for the hydropower non-linear function, and all the decomposed models are transformed into linear programming models that can be quickly solved. The operation simulation is carried out in a rolling manner according to the time sequence, and a simulation result correction mechanism is added to feedback and update the unreasonable simulation results layer by layer to ensure the solvability of the large-scale long-time series operation simulation model. The present invention provides simulation support for more reasonably determining the new energy capacity for complementary operation with hydropower stations.

[0007] Technical Solution of the Present Invention:

[0008] A multi-period nested optimization modeling and solving method for the complementary operation of hydropower, wind, and solar, the steps are as follows:

[0009] Step (1): Initial calculation conditions.

[0010] Historical observation data: including hourly natural runoff data for 12 consecutive years, hourly wind speed data at the hub height of the wind turbine, hourly solar radiation intensity data, and hourly air temperature data;

[0011] Basic data of hydropower stations: including installed capacity data, upper and lower limits of reservoir capacity, upper and lower limits of power generation flow, initial and final reservoir capacities, external transmission channel capacity, etc.;

[0012] Wind turbine characteristic data: including installed hub height, rated power, rated wind speed, cut-in wind speed, and cut-out wind speed;

[0013] Photovoltaic module data: including rated power, solar radiation intensity and photovoltaic panel temperature under standard test conditions, temperature coefficient, and photovoltaic panel temperature under normal working conditions.

[0014] Step (2): Construct a multi-year operation simulation model of hydropower, wind power, and photovoltaic power with a multi-year cycle and an annual step size.

[0015] Objective function 1: Minimize the sum of the deviations between the multi-year annual power generation of the hybrid hydropower, wind power, and photovoltaic power system and the mean value.

[0016]

[0017] In the formula: is the cycle of the multi-year operation simulation model; is the output of the hydropower station during time period; is the output of the photovoltaic power station during time period; is the output of the wind power station during time period.

[0018] Objective function 2: Maximize the total power generation of the hybrid hydropower, wind power, and photovoltaic power system over the years.

[0019]

[0020] In the formula: is the total number of hours in the time period.

[0021] The constraint conditions include: output function constraints of hydropower, wind power, and photovoltaic power stations, upper and lower limits of reservoir capacity, upper and lower limits of outflow discharge, upper and lower limits of power generation discharge, upper and lower limits of hydropower output, initial and final reservoir capacity constraints, water balance constraints, channel capacity constraints, etc. The mathematical expressions of the constraint condition models are as follows:

[0022] Hydropower station output expression:

[0023] Average reservoir capacity expression:

[0024] Photovoltaic power station output expression:

[0025] Wind power station output expression:

[0026] Water balance constraint:

[0027] Reservoir capacity constraint:

[0028] Outflow discharge constraint:

[0029] Power generation discharge constraint:

[0030] Hydropower station output constraint:

[0031] Initial and final storage capacity constraints:

[0032] Channel constraints:

[0033] Photovoltaic power output conversion formula:

[0034] Wind power output conversion formula:

[0035] In the above formula, is the hydropower station output function; is the average storage capacity of the hydropower station at time; is the hydropower station at time of the generation flow; is the hydropower station at time of the storage capacity; is the capacity of the photovoltaic power station; is the built photovoltaic power station at time of the output per unit installed capacity; is the capacity of the wind power station; is the built wind power station at time of the output per unit installed capacity; and are respectively the inflow and the spillage flow of the hydropower station at time; and are respectively the minimum allowable storage capacity and the maximum allowable storage capacity of the hydropower station at time; and are respectively the minimum allowable out - flow and the maximum allowable out - flow; and are respectively the minimum allowable generation flow and the maximum allowable generation flow of the hydropower station; and are respectively the minimum allowable output and the maximum allowable output of the hydropower station; and are the initial storage capacity and the final storage capacity of the hydropower station during the simulation period; is the maximum capacity of the hydropower station transmission channel; and are respectively the solar radiation intensity and the photovoltaic panel temperature under standard test conditions; Photovoltaic panel temperature under normal working conditions; is the temperature coefficient; 、 and are respectively the photovoltaic panel working temperature, the air temperature and the solar radiation intensity of the photovoltaic power station during period; , and are the cut-in wind speed, rated wind speed and cut-out wind speed of the wind turbine respectively; is the hub height wind speed of the wind farm during period.

[0036] Step (3): Construct a water-wind-solar annual operation simulation model with a year as the cycle and a month as the step.

[0037] Objective function 1: Minimize the sum of the deviations of the power generation of the water-wind-solar complementary system in different months from the average value.

[0038]

[0039] In the formula, , representing the cycle of the annual operation simulation.

[0040] Objective function 2: Maximize the total annual power generation of the water-wind-solar complementary system.

[0041]

[0042] In the formula, is the number of hours in the

[0043] The constraint conditions include the output function constraints of the water-wind-solar power stations, the upper and lower limits of the reservoir capacity, the upper and lower limits of the outflow flow rate, the upper and lower limits of the power generation flow rate, the upper and lower limits of the hydropower output, the initial and final reservoir capacity constraints, the water volume balance constraint, and the channel capacity constraint; except that the reservoir water volume balance constraint is replaced by the following formula, the other constraints are the same as those in step (2):

[0044]

[0045] If the annual operation simulation model cannot be solved, the boundary of the end-of-year reservoir capacity needs to be modified, and the initial and final reservoir capacity constraints are replaced by:

[0046]

[0047] Among them, is the percentage of the allowable reservoir capacity deviation.

[0048] In order to reduce the deviation between the end-of-year reservoir capacity correction amount and the end-of-year planned reservoir capacity, the allowable reservoir capacity deviation percentage is set in ascending order . The allowable reservoir capacity deviation percentage is set to 0.05%, 0.1%, 0.5%, 1%, 5%, 10%, 20%, 30%, 40% in sequence.

[0049] Step (4): Construct a water-wind-solar monthly operation simulation model with a month as the cycle and an hour as the step.

[0050] Objective function 1: The deviation between the reservoir storage at the end of the month and the storage boundary transmitted by the annual operation simulation model is minimized.

[0051]

[0052] Where: represents the simulation period of the monthly operation simulation.

[0053] Objective function 2: The total monthly power generation of the water-wind-solar complementary system is maximized.

[0054]

[0055] Where: , represents the number of hours in the

[0056] The constraint conditions include the output function constraints of the water-wind-solar power stations, the upper and lower limits of the reservoir storage, the upper and lower limits of the outflow discharge, the upper and lower limits of the power generation discharge, the upper and lower limits of the hydropower output, the initial and final reservoir storage constraints, the water balance constraints, and the channel capacity constraints; except that the water balance constraint of the reservoir is replaced by the following formula, the other constraints are the same as those in step (2):

[0057]

[0058] Convert the final reservoir storage constraint into an objective function, and retain the initial reservoir storage constraint:

[0059]

[0060] Step (5): Add the water-wind-solar complementary operation mode constraint to the monthly operation simulation model.

[0061] Through the analysis of the actual operation data, the water-wind-solar complementary operation mode constraint clusters and abstracts three water-wind-solar complementary operation modes: the double-peak mode (mainly occurring from January to May and from November to December); the single-peak mode (mostly occurring in June and October); the smooth mode (mostly occurring in July, August, and September). Specifically as follows:

[0062] ① Double-peak mode

[0063]

[0064] Where: is the maximum output of the complementary system on the th day; is the peak shaving parameter; represents the th year and the th month; That is The remainder when divided by 24 represents the time of day.

[0065] ② Unimodal mode

[0066]

[0067] ③ Smooth mode

[0068]

[0069] Step (6): Multi-time scale nested rolling correction operation simulation strategy based on the soft connection framework.

[0070] According to the coupling relationship between the hydropower simulation results at different time resolutions, without destroying the time continuity of the long-term hourly simulation results, the original continuous multi-year hourly resolution operation simulation is decomposed into a series of smaller scale annual, monthly, and hourly resolution models for rolling simulation. The upper layer transfers the operation boundary to the lower layer, and the lower layer feeds back the correction results to the upper layer.

[0071] The details are as follows: The multi-year operation simulation model conducts simulation with all remaining years as the cycle and years as the time step, and sends the annual reservoir storage capacity boundary of the hydropower station to the annual operation simulation model. The annual operation simulation model conducts simulation with years as the cycle and months as the time step; using the reservoir storage capacity boundary obtained from the multi-year operation simulation model as the input, it conducts rolling simulation starting from the initial period of the multi-year operation simulation model; then, it sends the monthly reservoir storage capacity boundary to the monthly operation simulation model. The monthly operation simulation model conducts simulation with hourly time steps and months as the cycle, using the reservoir storage capacity boundary obtained from the annual operation simulation model, and conducts rolling simulation starting from the first time period of the annual operation simulation model; the monthly-end reservoir storage capacity boundary obtained from the monthly operation simulation is fed back to the annual operation simulation once a month as the initial reservoir storage capacity for the remaining months of the year. That is to say, if the monthly operation simulation can meet the monthly reservoir storage capacity boundary of the hydropower station transmitted from the annual operation simulation model, it directly proceeds to the monthly operation simulation of the next month; otherwise, the corrected monthly-end reservoir storage capacity boundary will be returned to the annual operation simulation model of the current year as the starting reservoir storage capacity boundary of the current year, and the annual operation simulation of the remaining months of the current year will be re-run, and the corrected reservoir storage capacity boundaries of the remaining months will be re-transmitted to the monthly operation simulation model. Similarly, if the annual operation simulation of the remaining months cannot meet the annual reservoir storage capacity boundary obtained from the multi-year operation simulation model, the corrected annual-end reservoir storage capacity boundary will be returned to the multi-year operation simulation model as the initial reservoir storage capacity boundary of the remaining years, and the multi-year operation simulation will be re-run. The corrected reservoir storage capacity boundaries of the remaining years will be re-transmitted to the annual operation simulation model. It should be noted that if the monthly-end reservoir storage capacity is corrected in November, the corrected reservoir storage capacity will be returned to the annual operation simulation model, and there is no need to conduct annual operation simulation for the remaining months of the year; if it occurs in December, the corrected reservoir storage capacity will be directly fed back to the multi-year operation simulation model.

[0072] Step (7): Using the parallelogram two-dimensional interpolation method, without introducing 0-1 integer variables, linearly approximate and model the hydropower nonlinear output function, and convert all the decomposed models into linear programming models that can be quickly solved.

[0073] According to the historical actual operation data of the hydropower station, fit the nonlinear hydropower output function into the form of a two-variable quadratic polynomial.

[0074]

[0075] In the formula: is the two-variable quadratic function of the hydropower output function parameters.

[0076] Based on the parallelogram two-dimensional interpolation method, without introducing 0-1 integer variables, a linear approximation model is established for the hydroelectric nonlinear output function in the multi-year simulation model, annual simulation model, and monthly simulation model. The linearized expression is as follows:

[0077]

[0078]

[0079]

[0080]

[0081]

[0082]

[0083]

[0084]

[0085]

[0086]

[0087]

[0088] For clarity, the following introduction omits the subscripts . First, the hydroelectric output function is discretized into a non-orthogonal grid defined by parallelograms; represents the discrete gradient of the power generation flow rate, represents the vertices of the parallelogram, and are the sampling values of the power generation flow rate in the th column and th row, and the sampling values of the reservoir capacity in the th row, respectively, and . In addition and represent the vertex indices, , , and represent the number of division intervals in the horizontal direction and the number of division intervals in the vertical direction of the parallelogram network, respectively, and represent the linear auxiliary variables. The first formula determines that the distance between adjacent is equal. Therefore, in the sixth formula, represents the distance between adjacent and does not need to be represented by other parameters.

[0089] Step (8): With the help of the linear programming solver Gurobi, solve the simulation models of steps (2), (3), and (4) linearized by the parallelogram interpolation method in step (7) according to the operation simulation strategy proposed in step (6). Input the generated wind and photovoltaic installed capacities to be optimized into the simulation models of steps (2), (3), and (4), and solve to obtain the specific operation conditions such as hydropower output, photovoltaic output, wind power output, reservoir capacity change of the hydropower, wind, and solar complementary system, and hydropower station generation flow at hourly resolution for multiple years.

[0090] Advantages of the present invention: Without destroying the time continuity of the long-term hourly simulation results, based on the coupling relationship between hydropower simulation results at different time resolutions, the original continuous multi-year hourly simulation is decomposed into a series of smaller-scale annual, monthly, and hourly simulation models. Secondly, based on the parallelogram two-dimensional interpolation method, without introducing 0-1 integer variables, a linear approximation model is established for the hydropower nonlinear function, and all the decomposed models are transformed into linear programming models that can be quickly solved. The operation simulation is carried out in a rolling manner according to the time sequence, and a simulation result correction mechanism is added to feedback and update the unreasonable simulation results layer by layer to ensure the solvability of the large-scale long-time series operation simulation model. The present invention solves the challenge of large-scale nonlinear solution for multi-year operation simulation of hydropower, wind, and solar, and provides simulation support for more reasonably determining the new energy capacity for complementary operation with hydropower stations. Description of the Drawings

[0091] Figure 1 It is a schematic diagram of a multi-time scale nested rolling correction operation simulation strategy based on a soft connection framework;

[0092] Figure 2 It is a schematic diagram of a hydropower, wind, and solar complementary operation mode;

[0093] Figure 3 It is a schematic diagram of the distribution of sampling points for linearizing the power generation function;

[0094] Figure 4 It is a general solution framework diagram of the method of the present invention. Detailed Embodiment

[0095] The present invention will be further described below in conjunction with the drawings and embodiments.

[0096] The overall process of the present invention is as Figure 4 shown.

[0097] In this embodiment, a water-wind-solar clean energy base located at the Guangzhao Hydropower Station in the Beipanjiang River Basin, Guizhou Province, China is taken as an implementation case to test the present invention. The Guangzhao Hydropower Station is located in the middle reaches of the Beipanjiang River in Guizhou Province, China, and is the leading hydropower station in the Beipanjiang River Basin hydropower stations. The clean hydropower of the Guangzhao Hydropower Station is transmitted to Guangdong Province, China through the Xing'an DC connection line. To effectively utilize the wind and solar resources and fully utilize the flexibility of hydropower, a water-wind-solar complementary operation base is currently planned to be built at the Guangzhao Hydropower Station. The wind power station, photovoltaic power station and the existing hydropower station operate complementarily, and complete the complementary power generation plan as a hybrid power source. After grid connection, it is transmitted to the Guangdong power grid through the existing Xing'an DC connection line. Since the water-wind-solar complementary system is still under construction, there is still controversy about the optimal capacity of its wind power and photovoltaic power. Due to the significant interannual variations in wind and solar resources and runoff, it is necessary to conduct multi-year operation simulations of water-wind-solar to provide a simulation basis for the planning of wind and solar capacity. Therefore, in this embodiment, the water-wind-solar complementary operation system of the Guangzhao Hydropower Station is taken as a simulation case for multi-year continuous operation simulation to provide simulation support for the planning of wind and solar capacity.

[0098] Step (1): Initial calculation conditions.

[0099] Historical observation data: Collect the hourly natural runoff data of the Beipanjiang Guangzhao Hydropower Station for 12 consecutive years from 2008 to 2019. The data is sourced from the power grid company. Download the hourly wind speed data, solar radiation intensity data, and air temperature data of the Beipanjiang Guangzhao Hydropower Station for 12 consecutive years from 2008 to 2019 from the official website of the open-source dataset ERA5 of the European Centre for Medium-Range Weather Forecasts (Copernicus Climate DataStore of the European Centre for Medium-Range Weather Forecasts).

[0100] Basic data of the hydropower station: The relevant data of the Guangzhao Hydropower Station is sourced from the Beipanjiang River Basin Centralized Control Center, including installed capacity data, upper and lower reservoir capacity limits, upper and lower limits of power generation flow, initial and final reservoir capacities, external transmission channel capacities, etc.;

[0101] Fan characteristic data: including installed hub height, rated power, rated wind speed, cut-in wind speed, and cut-out wind speed, sourced from the official website of ERA5 (Copernicus Climate Data Store of the European Centre for Medium-Range Weather Forecasts);

[0102] Photovoltaic module data: including rated power, solar radiation intensity and photovoltaic panel temperature under standard test conditions, temperature coefficient, and photovoltaic panel temperature under normal operating conditions, sourced from the ERA5 official website (Copernicus Climate Data Store of the European Centre for Medium-Range Weather Forecasts).

[0103] Step (2): Construct a multi-year operation simulation model for water, wind, and light with a multi-year cycle and an annual step size.

[0104] Objective function 1: Minimize the sum of the deviations between the multi-year annual power generation of the water-wind-light complementary system and the mean.

[0105]

[0106] Wherein: is the cycle of the multi-year operation simulation model; is the output of the hydropower station at time period; is the output of the photovoltaic power station at time period; is the output of the wind power station at time period.

[0107] Objective function 2: Maximize the total power generation of the water-wind-light complementary system over the years.

[0108]

[0109] Wherein: is total number of hours in the time period.

[0110] The constraint conditions include: output function constraints of water, wind, and light power stations, upper and lower limits of reservoir capacity, upper and lower limits of outflow discharge, upper and lower limits of power generation discharge, upper and lower limits of hydropower output, initial and final reservoir capacity constraints, water balance constraints, channel capacity constraints, etc. The mathematical expressions of the constraint condition models are as follows:

[0111] Hydropower station output expression:

[0112] Average reservoir capacity expression:

[0113] Photovoltaic power station output expression:

[0114] Wind power station output expression:

[0115] Water balance constraint:

[0116] Reservoir capacity constraint:

[0117] Outflow rate constraint:

[0118] Power generation flow rate constraint:

[0119] Hydropower station output constraint:

[0120] Initial and final reservoir capacity constraints:

[0121] Channel constraint:

[0122] Photovoltaic output conversion formula:

[0123] Wind power output conversion formula:

[0124] In the above formula, is the hydropower station output function; is the average reservoir capacity of the hydropower station at time; is the power generation flow rate of the hydropower station at time; is the reservoir capacity of the hydropower station at time; is the capacity of the photovoltaic power station; is the output per unit installed capacity of the built photovoltaic power station at time; is the capacity of the wind power station; is the output per unit installed capacity of the built wind power station at time; and are the inflow rate and the discharged water flow rate of the hydropower station at time respectively; and are the minimum allowable reservoir capacity and the maximum allowable reservoir capacity of the hydropower station at time respectively; and are the minimum allowable outflow rate and the maximum allowable outflow rate respectively; and are the minimum allowable power generation flow rate and the maximum allowable power generation flow rate of the hydropower station respectively; and are the minimum allowable output and the maximum allowable output of the hydropower station respectively; and are the initial reservoir capacity and the final reservoir capacity of the hydropower station during the simulation period; is the maximum capacity of the hydropower station transmission channel; and are the solar radiation intensity and the photovoltaic panel temperature under standard test conditions, respectively; the photovoltaic panel temperature under normal operating conditions; is the temperature coefficient; 、 and are the working temperature of the photovoltaic panel, the air temperature and the solar radiation intensity of the photovoltaic power station during period, respectively; 、 and are the cut-in wind speed, the rated wind speed and the cut-out wind speed of the wind turbine, respectively; is the wind speed at the hub height of the wind power station during period.

[0125] Step (3): Construct an annual operation simulation model of hydropower, wind power and photovoltaic with a monthly step.

[0126] Objective function 1: Minimize the sum of the deviations between the monthly power generations of the hydropower, wind power and photovoltaic complementary system and the average value.

[0127]

[0128] In the formula, represents the period of the annual operation simulation.

[0129] Objective function 2: Maximize the total annual power generation of the hydropower, wind power and photovoltaic complementary system.

[0130]

[0131] In the formula, is the number of hours during the period.

[0132] The constraint conditions include the output function constraints of hydropower, wind power and photovoltaic power stations, the upper and lower limits of reservoir capacity, the upper and lower limits of the outflow discharge, the upper and lower limits of the power generation discharge, the upper and lower limits of hydropower output, the initial and final reservoir capacity constraints, the water volume balance constraint, and the channel capacity constraint; except that the reservoir water volume balance constraint is replaced by the following formula, the other constraints are the same as those in step (2):

[0133]

[0134] If the annual operation simulation model cannot be solved, it is necessary to modify the boundary of the end-of-year reservoir capacity and replace the initial and final reservoir capacity constraints with:

[0135]

[0136] Among them, is the percentage of the allowable reservoir capacity deviation.

[0137] To reduce the deviation between the end-of-year reservoir capacity correction and the end-of-year planned reservoir capacity, the allowable reservoir capacity deviation percentage is set in ascending order. The allowable reservoir capacity deviation percentage is successively set to 0.05%, 0.1%, 0.5%, 1%, 5%, 10%, 20%, 30%, and 40%.

[0138] Step (4): Construct a monthly operation simulation model of water, wind, and light with an hourly step.

[0139] Objective function 1: Minimize the deviation between the end-of-month reservoir capacity and the reservoir capacity boundary transmitted by the annual operation simulation model.

[0140]

[0141] In the formula: represents the simulation period of the monthly operation simulation.

[0142] Objective function 2: Maximize the total monthly power generation of the water, wind, and light complementary system.

[0143]

[0144] In the formula: , represents the number of hours in the time period.

[0145] The constraint conditions include the output function constraints of the water, wind, and light power stations, the upper and lower limits of the reservoir capacity, the upper and lower limits of the outflow flow rate, the upper and lower limits of the power generation flow rate, the upper and lower limits of the hydropower output, the initial and final reservoir capacity constraints, the water volume balance constraint, and the channel capacity constraint; except that the water volume balance constraint of the reservoir is replaced by the following formula, the other constraints are the same as those in step (2):

[0146]

[0147] Convert the final reservoir capacity constraint into an objective function and retain the initial reservoir capacity constraint:

[0148]

[0149] Step (5): Add the constraints of the water, wind, and light complementary operation mode to the monthly operation simulation model.

[0150] The constraints of the water, wind, and light complementary operation mode are obtained by analyzing the actual operation data and clustering and abstracting the three water, wind, and light complementary operation modes: the double-peak mode (mainly occurring from January to May and from November to December); the single-peak mode (mostly occurring in June and October); the smooth mode (mostly occurring in July, August, and September).

[0151] The schematic diagram of the water, wind, and light complementary operation mode is as Figure 2 shown.

[0152] ① Double-peak mode

[0153]

[0154] Wherein: is the maximum output of the complementary system on the th day; is the peak regulation parameter; represents the th year and the th month; That is the remainder of divided by 24, representing the time of day. For example if

[0155] ② Single-peak mode

[0156]

[0157] ③ Smooth mode

[0158]

[0159] Step (6): Multi-time scale nested rolling correction operation simulation strategy based on the soft connection framework.

[0160] According to the coupling relationship between the hydropower simulation results at different time resolutions, without destroying the time continuity of the long-term hourly simulation results, the original continuous multi-year hourly resolution operation simulation is decomposed into a series of smaller scale annual, monthly, and hourly resolution models for rolling simulation. The upper layer transfers the operation boundary to the lower layer, and the lower layer feeds back the correction results to the upper layer. As Figure 1 shown.

[0161] The specific steps are as follows: The multi-year operation simulation model conducts simulation with all remaining years as the cycle and years as the time step, and sends the annual reservoir storage capacity boundary of the hydropower station to the annual operation simulation model. The annual operation simulation model conducts simulation with years as the cycle and months as the time step; it uses the reservoir storage capacity boundary obtained from the multi-year operation simulation model as the input and starts rolling simulation from the initial period of the multi-year operation simulation model; then, it sends the monthly reservoir storage capacity boundary to the monthly operation simulation model. The monthly operation simulation model conducts simulation with hourly time steps and months as the cycle, uses the reservoir storage capacity boundary obtained from the annual operation simulation model, and starts rolling simulation from the first time period of the annual operation simulation model; the monthly reservoir storage capacity boundary obtained from the monthly operation simulation is fed back to the annual operation simulation once a month as the initial reservoir storage capacity for the remaining months of the year. That is to say, if the monthly operation simulation can meet the monthly reservoir storage capacity boundary of the hydropower station transmitted from the annual operation simulation model, the monthly operation simulation for the next month will be directly carried out; otherwise, the corrected monthly-end reservoir storage capacity boundary will be returned to the annual operation simulation model of the current year as the starting reservoir storage capacity boundary of the current year, and the annual operation simulation for the remaining months of the current year will be re-run, and the corrected reservoir storage capacity boundaries for the remaining months will be re-transmitted to the monthly operation simulation model. Similarly, if the annual operation simulation for the remaining months cannot meet the annual reservoir storage capacity boundary obtained from the multi-year operation simulation model, the corrected annual-end reservoir storage capacity boundary will be returned to the multi-year operation simulation model as the initial reservoir storage capacity boundary for the remaining years, and the multi-year operation simulation will be re-run. The corrected reservoir storage capacity boundaries for the remaining years will be re-transmitted to the annual operation simulation model. It should be noted that if the monthly-end reservoir storage capacity is corrected in November, the corrected reservoir storage capacity will be returned to the annual operation simulation model, and there is no need to conduct annual operation simulation for the remaining months of the year; if it occurs in December, the corrected reservoir storage capacity will be directly fed back to the multi-year operation simulation model.

[0162] Step (7): Adopt the parallelogram two-dimensional interpolation method to linearly approximate and model the hydropower nonlinear output function without introducing 0-1 integer variables, and convert all the decomposed models into linear programming models that can be quickly solved.

[0163] According to the historical actual operation data of the hydropower station, the nonlinear hydropower output function is fitted into the form of a two-variable quadratic polynomial.

[0164]

[0165] In the formula: is the two-variable quadratic function of the hydropower output function parameters.

[0166] Based on the parallelogram two-dimensional interpolation method, without introducing 0-1 integer variables, a linear approximation model is established for the hydroelectric non-linear output function in the multi-year simulation model, annual simulation model, and monthly simulation model. The linearized expression is as follows:

[0167]

[0168]

[0169]

[0170]

[0171]

[0172]

[0173]

[0174]

[0175]

[0176]

[0177]

[0178] For clarity, the following introduction omits the subscripts . First, the hydroelectric output function is discretized into a non-orthogonal grid defined by parallelograms; represents the discrete gradient of the power generation flow rate, represents the vertices of the parallelogram, and are the sampling values of the power generation flow rate in the th column and th row, and the sampling values of the reservoir capacity in the th row, respectively, and . In addition and represent the vertex indices, , , and represent the number of division intervals in the horizontal direction ( = 4) and the number of division intervals in the vertical direction ( = 3) of the parallelogram network, respectively, and represent linear auxiliary variables. The first formula determines that the distance between adjacent is equal, so in the sixth formula represents the adjacent The distance between them needs no other parameters to represent. In this example, the distribution of linearized sampling points of the power generation function is as Figure 3 shown, where , .

[0179] Step (8): With the help of the linear programming solver Gurobi, solve the linear programming model after parallelogram interpolation linearization in steps (2), (3), and (4) according to the operation simulation strategy proposed in step (6). Randomly generate 10 groups of combinations of photovoltaic capacity and wind power capacity within the allowable range, and input them into the continuous 12-year hourly water-wind-solar operation simulation model after variable time-scale decomposition respectively, and record the solution results and solution time. For comparison, establish a continuous 12-year hourly traditional water-wind-solar operation simulation model without variable time-scale decomposition, and explore the performance of the operation simulation technology proposed in the present invention in terms of solution accuracy and efficiency. The continuous 12-year hourly traditional operation simulation model without variable time-scale decomposition takes the maximum total power generation of water-wind-solar over the years as the objective function, and at the same time sets the relevant constraint conditions in the previous steps, inputs the same wind-solar resources and runoff data as in the previous steps, and also conducts parallelogram interpolation linear approximation modeling on the hydropower non-linear function. Input the above 10 groups of combinations of photovoltaic capacity and wind power capacity into the traditional model respectively for continuous 12-year hourly water-wind-solar complementary operation simulation, and record the solution results and solution time. The solution results are shown in Table 1.

[0180] Table 1 Comparison of solution results between the decomposed model (the present invention) and the undecomposed model

[0181]

[0182] As can be seen from Table 1, compared with the continuous 12-year hourly traditional water-wind-solar operation simulation model without variable time-scale decomposition, the continuous 12-year hourly water-wind-solar operation simulation model after variable time-scale decomposition can reduce the average solution duration from 4103 s to 616 s. Since the new energy capacity planning of the basin water-wind-solar complementary operation base needs to conduct multiple multi-year operation simulation simulations under different assumed scenarios of new energy capacity, the advantage of this operation simulation model in terms of solution efficiency will be significantly expanded. At the same time, the average similarity of the solution results of the two is as high as 99.66%. Therefore, the water-wind-solar multi-energy complementary multi-cycle nested optimization modeling and solution technology proposed in the present invention realizes a significant improvement in solution efficiency while taking into account the solution accuracy.

Claims

1. A multi-period nested optimization modeling and solving method for water, wind and solar multi-energy complementarity, characterized in that: Here are the steps: Step (1): initial calculation conditions; Including historical observation data, basic data of hydropower stations, wind turbine characteristic data and photovoltaic module data; Step (2): construct a multi-year operation simulation model for water, wind and solar power with a multi-year cycle and a year as the step length; Objective function 1: Minimize the sum of the deviations between the annual power generation of the hydro-wind-solar hybrid system and the mean value; Where: It is a cycle of running simulation models over many years; The hydroelectric station is Output during the time period; The photovoltaic power station is Output during the time period; The wind farm is Output during the time period; Objective function 2: Maximize the total power generation of the water-wind-solar hybrid system over many years; Where: yes Total number of hours in the session; The constraints include: output function constraints of hydropower stations, wind power stations, and photovoltaic stations, upper and lower limits of storage capacity, upper and lower limits of outflow, upper and lower limits of power generation flow, upper and lower limits of hydropower output, initial and final storage capacity constraints, water balance constraints, and channel capacity constraints; the mathematical expression of the constraint model is as follows: Hydropower station output expression: Average storage capacity expression: Photovoltaic power station output expression: Wind power station output expression: Water balance constraints: Storage capacity constraints: Outbound traffic constraints: Power generation flow constraints: Hydropower station output constraints: Initial and final storage capacity constraints: Channel constraints: Photovoltaic output conversion formula: Wind power output conversion formula: In the above formula, is the hydropower station output function; The hydroelectric station is Average storage capacity at the time; The hydroelectric station is The power generation flow at the moment; The hydroelectric station is Storage capacity at a given moment; is the capacity of the PV plant; The photovoltaic power station has been built in Output per installed capacity at any moment; is the capacity of the wind farm; Is the wind power station built in Output per installed capacity at any moment; and The hydropower stations are Inflow and outflow at each time; and The hydropower stations are The minimum and maximum allowable storage capacity at the time; and They are the minimum allowed outbound flow and the maximum allowed outbound flow; and They are the minimum and maximum permissible power generation flow of the hydropower station; and They are the minimum and maximum allowable output of the hydropower station; and are the initial and final storage capacities of the hydropower station during the simulation period; It is the maximum capacity of the transmission channel of the hydropower station; and are the solar radiation intensity and photovoltaic panel temperature under standard test conditions, respectively; PV panel temperature under normal operating conditions; is the temperature coefficient; , and The photovoltaic power station is PV panel operating temperature, air temperature and solar radiation intensity during the period; , and They are the cut-in wind speed, rated wind speed and cut-out wind speed of the fan respectively; For wind power plants Hub height wind speed for the time period; Step (3): construct a water, wind and light year operation simulation model with a year as the cycle and a month as the step length; Objective function 1: Minimize the sum of the deviations between the power generation of the hydro-wind-solar hybrid system in different months and the average value; In the formula, , represents the cycle of annual operation simulation; Objective function 2: Maximize the total annual power generation of the water-wind-solar complementary system; In the formula, yes The number of hours in the session; Among the constraints, except that the reservoir water balance constraint is replaced by the following formula, the remaining constraints are the same as those in step (2): If the annual operation simulation model cannot be solved, it is necessary to modify the reservoir capacity boundary at the end of the year and replace the initial and final capacity constraints with: in, is the percentage of allowable storage capacity deviation; Step (4): construct a water, wind and light monthly operation simulation model with a monthly cycle and an hourly step length; Objective function 1: The deviation between the reservoir capacity at the end of the month and the reservoir capacity boundary delivered by the annual operation simulation model is minimized; Where: represents the simulation period of the monthly operation simulation; Objective function 2: The total monthly power generation of the water-wind-solar hybrid system is maximized; Where: ,represent The number of hours in the session; Among the constraints, except that the reservoir water balance constraint is replaced by the following formula, the remaining constraints are the same as those in step (2): The final reservoir capacity constraint is transformed into the objective function, and the initial reservoir capacity constraint is retained: Step (5): adding the water-wind-solar complementary operation mode constraints to the monthly operation simulation model; Through the analysis of actual operation data, three water-wind-solar complementary operation modes are clustered and abstracted: bimodal mode, unimodal mode and smooth mode; the details are as follows: ① Bimodal model Where: The complementary system is The maximum output of the sky; is the peak regulation parameter; Representative Year The number of days in the month; Right now The remainder when divided by 24 represents the time of day; ②Single peak mode ③Smooth mode Step (6): Run simulation strategy based on multi-time scale nested rolling correction of soft connection framework; According to the coupling relationship between the hydropower simulation results at different time resolutions, without destroying the temporal continuity of the long-term hourly simulation results, the original continuous multi-year hourly resolution operation simulation is decomposed into a series of smaller-scale annual, monthly and hourly resolution models for rolling simulation. The upper layer transfers the operation boundary to the lower layer, and the lower layer feeds back the correction results to the upper layer. Step (7): using the parallelogram two-dimensional interpolation method, without introducing 0-1 integer variables, linear approximation modeling is performed on the hydropower nonlinear output function, and all decomposed models are converted into fast-solving linear programming models; According to the historical actual operation data of the hydropower station, the nonlinear hydropower station output function is fitted into the form of a two-variable quadratic polynomial; Where: It is a bivariate quadratic function of the hydropower station output function Parameters; Based on the parallelogram two-dimensional interpolation method, without introducing 0-1 integer variables, the nonlinear output function of hydropower in the multi-year simulation model, annual simulation model, and monthly simulation model is linearly approximated and modeled; the linearized expression is as follows: Firstly, the hydropower output function is discretized into a non-orthogonal grid defined by parallelograms; represents the discrete gradient of the power generation flow, represents the vertices of the parallelogram, and The power generation flow is Ledi The sampling value and storage capacity of the row The sample value of the row, and ;also and represents the vertex index, , , and They represent the number of horizontal and vertical divisions of the parallelogram network, respectively. and represents linear auxiliary variables; Step (8): With the help of the linear programming solver Gurobi, the simulation model of steps (2)(3)(4) after linearization by the parallelogram interpolation method in step (7) is solved according to the operation simulation strategy proposed in step (6), and the generated wind power and photovoltaic installed capacity to be optimized are input into the simulation model of steps (2)(3)(4), so as to obtain the specific operation status of the hydropower output, photovoltaic output, wind power output, hydropower station reservoir capacity change, and hydropower station power generation flow of the water-wind-solar complementary system with multi-year hourly resolution.

2. According to claim 1, a multi-period nested optimization modeling solution method for water-wind-solar multi-energy complementarity is characterized in that: In step (1): Historical observation data: including hourly natural runoff data for 12 consecutive years, hourly wind speed data at the height of the wind turbine hub, hourly solar radiation intensity data, and hourly air temperature data; Basic data of hydropower stations: including installed capacity data, upper and lower limits of reservoir capacity, upper and lower limits of power generation flow, initial and final reservoir capacity, and transmission channel capacity; Wind turbine characteristic data: including installation hub height, rated power, rated wind speed, cut-in wind speed and cut-out wind speed; PV module data: including rated power, solar radiation intensity and PV panel temperature under standard test conditions, temperature coefficient and PV panel temperature under normal working conditions.

3. The method for solving the multi-period nested optimization modeling of water-wind-solar multi-energy complementarity according to claim 1 is characterized in that: Step (6) is as follows: the multi-year operation simulation model uses all remaining years as a cycle and a year as a step length to perform simulation, and sends the annual reservoir capacity boundary of the hydropower station to the annual operation simulation model; the annual operation simulation model uses a year as a cycle and a month as a step length to perform simulation; using the reservoir capacity boundary obtained by the multi-year operation simulation model as input, a rolling simulation is started from the initial period of the multi-year operation simulation model; then, the monthly reservoir capacity boundary is sent to the monthly operation simulation model; the monthly operation simulation model uses an hourly step length and a month as a cycle, and uses the reservoir capacity boundary obtained by the annual operation simulation model to start a rolling simulation from the first time period of the annual operation simulation model; the reservoir capacity boundary at the end of the month obtained by the monthly operation simulation is fed back to the annual operation simulation once a month as the initial reservoir capacity for the remaining months of the year; if the monthly operation simulation can meet the monthly reservoir capacity boundary of the hydropower station transmitted from the annual operation simulation model, then Directly carry out the monthly operation simulation for the next month; otherwise, the revised reservoir capacity boundary at the end of the month will be returned to the annual operation simulation model of the current year as the starting reservoir capacity boundary for the current year, and the annual operation simulation for the remaining months of the current year will be re-run, and the revised reservoir capacity boundary for the remaining months will be re-transferred to the monthly operation simulation model; similarly, if the annual operation simulation for the remaining months cannot satisfy the annual reservoir capacity boundary obtained from the multi-year operation simulation model, the revised reservoir capacity boundary at the end of the year will be returned to the multi-year operation simulation model as the initial reservoir capacity boundary for the remaining years, and the multi-year operation simulation will be re-run; the corrected reservoir capacity boundary for the remaining years will be re-transferred to the annual operation simulation model; it should be noted that if the end-of-month reservoir capacity correction is carried out in November, the revised reservoir capacity will be returned to the annual operation simulation model, and there is no need to carry out annual operation simulation for the remaining months of the year; If it occurs in December, the revised reservoir capacity will be directly fed back into the multi-year operational simulation model.

Images