A firelight storage system scheduling method for regional power grid photovoltaic

Abstract

The application discloses a fire-light storage system scheduling method for regional power grid photovoltaic, relates to the technical field of power system scheduling, and has the technical scheme as follows: data collection and prediction, establishment of a main equipment model of a power system, establishment of a fire-light storage peak shaving hierarchical scheduling model including an upper layer model and a lower layer model, determination of an objective function, setting of a constraint condition, determination of a solution process of the fire-light storage peak shaving hierarchical scheduling model, solving of optimal scheduling of unit start-stop peak shaving stages by the upper layer model with the minimum equivalent net load fluctuation of the power system as a target, solving of output scheduling of each unit in the lower layer model with the minimum operation cost of the power system scheduling period as a target, linearization processing of functions inconvenient to bring into the fire-light storage peak shaving hierarchical scheduling model for solving, and finally simulation test in a real scene as a case; and the effect is that the consumption level of the regional power grid for photovoltaic is improved, and the operation cost of the power system is reduced.

Description

Technical Field

[0001] This invention relates to the field of power system dispatching technology, and more specifically, to a method for dispatching a solar-thermal-storage power grid system for regional power grid photovoltaics. Background Technology

[0002] With the rapid development of renewable energy, the proportion of photovoltaic power generation in the power grid is constantly increasing. However, the intermittency and uncertainty of its output pose a huge challenge to the stable operation and peak-shaving capacity of the power grid. Although traditional thermal power units have strong regulation capabilities, their operating costs are high and they pose environmental emission problems. The introduction of thermal-solar-storage systems provides a new technical means to balance the fluctuations in photovoltaic output. Therefore, researching a thermal-solar-storage system scheduling method for regional power grid photovoltaics is of great significance for improving the overall operating efficiency and stability of the power grid. Summary of the Invention

[0003] To address the shortcomings of existing technologies, the present invention aims to provide a method for scheduling a thermal-solar-storage system for regional power grid photovoltaic applications.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A method for scheduling a thermal-solar-storage system for regional power grid photovoltaic systems includes the following steps:

[0006] Collect photovoltaic power forecast and load data of regional power grid, determine the configuration parameters of power system, and obtain the start-up characteristics of the main equipment of thermal power plants and units implementing gas-steam combined cycle power generation during the start-up, shutdown and peak shaving phases.

[0007] Establish models of the main equipment in the power system, including a thermal power unit output power model and an energy storage power station model.

[0008] A hierarchical scheduling model for fire storage and peak shaving is established, including an upper-level model and a lower-level model. The objective functions of the upper-level model and the lower-level model are determined, and the constraints of the upper-level model and the lower-level model are set.

[0009] The solution process of the hierarchical scheduling model for thermal power generation and energy storage peak shaving is determined. The upper-level model aims to minimize the equivalent net load fluctuation of the power system and solve for the optimal scheduling during the start-up and shutdown peak shaving phase of the units. The lower-level model aims to minimize the operating cost within the power system scheduling cycle and solve for the output scheduling of each unit. Functions that are inconvenient to be substituted into the solution of the hierarchical scheduling model for thermal power generation and energy storage peak shaving are linearized.

[0010] The obtained scheduling scheme is implemented in the regional power grid and simulated. The system's operating status is monitored in real time, and adjustments and optimizations are made according to the actual situation.

[0011] Preferably, the process involves collecting photovoltaic power forecasts and load data from the regional power grid, determining power system configuration parameters, and obtaining the start-up characteristics of major equipment in thermal power plants and units implementing gas-steam combined cycle power generation during start-up, shutdown, and peak shaving phases, including:

[0012] Record typical daily photovoltaic and load conditions of the regional power grid;

[0013] The power system configuration parameters include 220MW of photovoltaic capacity, 44MW / 88MWh of energy storage power station capacity, and a system spinning reserve capacity of 10% of the load.

[0014] The thermal power plant is mainly equipped with three Type I gas turbines and one Type II gas turbine.

[0015] Thermal power plants implement gas-steam combined cycle power generation. For units in the start-up, shutdown, and peak-shaving phases, the start-up characteristics of gas-steam combined cycle power generation in actual operation can be divided into three processes: gas turbine start-up, grid connection and power generation to reach maximum load; waste heat boiler start-up and pipe warm-up; steam turbine start-up and grid connection and power generation to reach maximum load; when the unit is shut down, it is shut down in the order of steam turbine-waste heat boiler-gas turbine.

[0016] Preferably, a model of the main equipment of the power system is established, which includes a model of the output power of thermal power units and a model of energy storage power stations;

[0017] Thermal power unit output power model:

[0018] For the number is j The units, during the regular peak-shaving phase i Effort output at any given moment can be expressed as:

[0019] (1)

[0020] In the formula: p gt,j,i —— Gas turbine power; r i —Combined cycle steam-fuel ratio;

[0021] Based on actual operating data of thermal power plants, the output during the start-up and shutdown phases of the units can be represented by the following piecewise functions (2) and (3):

[0022] (2)

[0023] (3)

[0024] In the formula, t —The time elapsed during the unit start-up and shutdown phases; Pgt,j,max —The rated maximum power of the gas turbine; P gen,j,max —The maximum power unit of the combined cycle unit, ; , —Climb rate during gas turbine startup and shutdown; —Maximum permissible output power; , — The ramp rate during turbine startup and shutdown; i a , i b , i c , i d —Time points during the unit startup phase i e , i f , i g —Shutdown points for steam turbines, waste heat boilers, and gas turbines;

[0025] Energy storage power station model:

[0026] Energy storage power stations i The state of charge at time t is shown in the following equation:

[0027] (4)

[0028] In the formula, S i —— i The state of charge of the energy storage power station at all times; —Data collection interval; E bn —Rated capacity of the energy storage power station; P bat,cha,i-1 , P bat,dis,i-1 —The charging and discharging power of the energy storage station at the previous moment; h bat,cha , h bat,dis — The charging and discharging efficiency of energy storage power stations.

[0029] Preferably, the hierarchical scheduling model for peak shaving by thermal power, solar power, and energy storage includes an upper-level model and a lower-level model. The objective functions of the upper-level and lower-level models are determined, and the constraints of the upper-level and lower-level models are set, including:

[0030] Based on photovoltaic day-ahead power forecasting and the dynamic model of start-up and shutdown peak-shaving units, an upper-level optimization model is established to optimize the downtime and downtime duration of start-up and shutdown peak-shaving units with the objective of minimizing the standard deviation of the equivalent net load fluctuation after adjustment by the start-up and shutdown peak-shaving units; for units numbered as j , x The system is constantly shut down, and the downtime is [duration]. y The unit, —Maximum downward adjustment rate, —Maximum upward adjustment rate, —The adjusted output of the j-th thermal power unit at time i. —The downward ramp characteristic function of the i-th thermal power unit, and its output sequence can be temporarily represented as:

[0031] (5)

[0032] In the formula, i down , i down,end , i up , i up,en —The time points of the unit shutdown and startup process are shown in equation (6):

[0033] (6)

[0034] Objective function of the upper-level model:

[0035] (7)

[0036] In the formula, N T — Scheduling cycle; P load,i —— i System load at all times; P PV,max,i —— i Predicted photovoltaic power at any given time; P gen,j,i —— i Start and stop peak-shaving units at any time contribution;

[0037] The constraints of the upper-level model are as follows:

[0038] Considering that the unit needs to complete the restart operation within the remaining time of the scheduling cycle after shutdown, the constraints on the unit's shutdown time and duration are as follows:

[0039] (8)

[0040] A multi-energy complementary optimal scheduling model based on thermal, solar, and storage energy is established in the lower-level model. The optimization objective is to achieve the optimal economic operation of the power system within the scheduling cycle. The objective function is: (9)

[0041] In the formula, C gen —The cost of generating electricity from thermal power plants; C bat —Operating costs of energy storage power stations; C PV —Penalties for curtailment of solar power; C net —The cost of purchasing electricity from external networks;

[0042] The power generation cost of a thermal power plant mainly considers the unit's fuel cost and start-up and shutdown costs, as shown in equations (10) and (11) below; for the first type of unit, since its power regulation range is very small, the fuel cost can be simplified to a linear function of power:

[0043] (10)

[0044] (11)

[0045] In the formula, N gen —Number of generating units; C gas,j,i —Fuel costs of the generating unit; d gen —Power generation cost of Type I generating units; a, b, c —Fuel consumption coefficient; d gas —Local natural gas price; n ss —Number of unit start-ups and shutdowns; d ss —Cost of a single start-up and shutdown of the unit;

[0046] The operating cost of an energy storage power station is:

[0047] (12)

[0048] In the formula, d bat —The charging and discharging cost coefficient of energy storage;

[0049] The penalty for curtailment of solar power is:

[0050] (13)

[0051] In the formula, P PV,loss,i —— iThe amount of solar power curtailed at any given time; d PV —Levelized cost of electricity (LCOE) over the entire lifecycle of a photovoltaic power plant;

[0052] Cost of purchasing electricity from external networks:

[0053] (14)

[0054] In the formula, P net,i —— i Power purchased from the external network at any given time; d net , i —— i The State Grid electricity price is based on time-of-use pricing.

[0055] The constraints of the lower-level model include:

[0056] Neglecting system network losses, the sum of the power output from thermal power, solar power, and energy storage, and the purchased power, equals the real-time load; power system power balance constraints:

[0057] (15)

[0058] In the formula, P net,i —— i Electricity purchase via the national online shopping platform; P pv,loss,i for i Real-time solar curtailment volume;

[0059] For units in the start-up and shutdown peak shaving phase, their output power is consistent with equation (5); for units in the rotating peak shaving phase, their output power is limited by the unit's ramp rate and the upper limit of the rated maximum output. This represents the maximum permissible deviation of the j-th thermal power unit. This represents the minimum permissible deviation of the j-th thermal power unit; dynamic constraints of thermal power plant units:

[0060] (16)

[0061] This indicates the total number of thermal power units and the standby capacity constraints of thermal power plants.

[0062] (17)

[0063] In the formula, U gen,j,i --unit j exist i The state variable at that time, 0 represents the unit being shut down, and 1 represents the unit being running; Ri --system i Reserve capacity at any given time;

[0064] Energy storage power station charging and discharging power constraints:

[0065] (18)

[0066] In the formula, P bat,cha,max 、P bat,dis,max —Maximum charging and discharging power of the energy storage power station, in MW; U cha,i , U dis,i —A Boolean variable indicating the charging and discharging of an energy storage power station, where 0 indicates that it is not in a charging or discharging state, and 1 indicates that it is in a charging or discharging state;

[0067] State of charge constraints for energy storage power stations:

[0068] (19)

[0069] In the formula, S max 、S min —The upper and lower limits of the state of charge of energy storage power stations; S 1 , S NT —The state of charge at the beginning and end of the scheduling cycle;

[0070] External electricity purchase restrictions:

[0071] (20)

[0072] In the formula, P net,max —The upper limit of the power capacity for purchasing electricity from external networks.

[0073] Preferably, the solution process for the hierarchical scheduling model for thermal power generation and energy storage peak shaving is determined. The upper-level model aims to minimize the equivalent net load fluctuation of the power system to solve for the optimal scheduling during the start-up and shutdown peak shaving phases of the units. The lower-level model aims to minimize the operating cost within the power system scheduling cycle to solve for the output scheduling of each unit. Functions that are inconvenient to be incorporated into the solution of the hierarchical scheduling model for thermal power generation and energy storage peak shaving are linearized, including:

[0074] The upper-level model first obtains the net load characteristics through the photovoltaic day-ahead power prediction model, which serves as the input to the upper-level model; then, a combination algorithm of FCM and LSTM is used to achieve short-term prediction of photovoltaic power.

[0075] Based on the number of start-up and shutdown peak-shaving units, the particle swarm optimization algorithm is used to minimize the standard deviation of the equivalent net load fluctuation of the power system after unit regulation. The shutdown time and duration of each start-up and shutdown unit are solved iteratively. The equivalent net load after regulation by the previous start-up and shutdown peak-shaving unit is used as the input for the next cycle. With the goal of optimizing the operating economy within the power system scheduling cycle, the output scheduling of each unit is solved by calling the CPLEX solver through Matlab.

[0076] The following functions that are inconvenient to be incorporated into the solution of the fire-storage peak-shaving hierarchical scheduling model are linearized, including:

[0077] Unit fuel costs:

[0078] First of all C gas,j,i Convert to n segmental linear functions f (P gt,j,i ), the dividing points are Introducing variables w k,j,i Will P gt,j,i ,and f ( P gt,j,i ) is represented by equation (21). —The state identifier variable of the j-th thermal power unit in the k-th segment under the i-th operating state. —The state identifier variable of the j-th thermal power unit in the (r-1)-th segment under the i-th operating state. —The state identifier variable of the j-th thermal power unit in the first segment under the i-th operating state. —The state identifier variable of the j-th thermal power unit in the r-th segment under the i-th operating state. —The two-state identifier variable of the j-th thermal power unit in the n-th segment under the i-th operating state. This represents the output power of the i-th unit under the j-th type of equipment during the first scheduling period. This represents the output power of the i-th unit under the j-th type of equipment during the r-th scheduling period. Let represent the output power of the i-th unit under the j-th type of equipment during the (n+1)-th scheduling period. The constraints of each variable are shown in equation (22):

[0079] (twenty one)

[0080] (twenty two)

[0081] Energy storage power station charging and discharging power constraints:

[0082] The constraints are transformed using the Big-M method:

[0083] (twenty three)

[0084] In the formula, M is a sufficiently large constant.

[0085] Preferably, the solved scheduling scheme is implemented in the regional power grid and simulated, the operating status of the power system is monitored in real time, and adjustments and optimizations are made according to the actual situation, including:

[0086] Based on the current photovoltaic forecast and peak-shaving unit start-up and shutdown model, the particle swarm optimization algorithm is used to solve the operation scheduling of start-up and shutdown units of thermal power plants according to the number of start-up and shutdown peak-shaving units put into operation; after putting into operation different numbers of start-up and shutdown peak-shaving units, the equivalent net load result of the power system is solved.

[0087] Based on the number of units put into operation for peak shaving and shutdown, the upper-level model can generate three optimal scheduling schemes. Based on different scheduling schemes, four scheduling scenarios are formed. Simulations are performed on these four scheduling scenarios to obtain the operating cost composition of the power system under various scheduling scenarios.

[0088] Compared with the prior art, the present invention has the following beneficial effects:

[0089] It improves the stability of the regional power grid. Through the rational scheduling of the thermal-solar-storage system, it can effectively balance the intermittency and volatility of photovoltaic power generation and ensure the stability of the grid voltage and frequency.

[0090] It has improved the absorption capacity of renewable energy, prioritized the use of photovoltaic power generation to meet load demand, and stored excess electricity through energy storage systems, thus reducing the phenomenon of curtailment of solar power.

[0091] Operating costs were reduced by optimizing scheduling strategies and rationally allocating the output of thermal power generation and energy storage systems, thereby reducing fuel consumption and equipment wear.

[0092] It has good adaptability and scalability, and can be adjusted and optimized according to the characteristics and needs of different regional power grids. Attached Figure Description

[0093] Figure 1 The present invention provides a flowchart of a scheduling method for a thermal-solar-storage system oriented towards regional power grid photovoltaics;

[0094] Figure 2 The flowchart for solving the hierarchical peak-shaving scheduling model of the thermal-solar-storage system proposed in this invention is as follows: A method for scheduling regional power grid photovoltaic systems.

[0095] Figure 3 This invention proposes a method for scheduling a thermal-solar-storage system for regional power grid photovoltaic systems, which includes the startup characteristics of a gas-steam combined cycle unit.

[0096] Figure 4 This invention proposes a scheduling method for a solar-thermal-storage system for regional power grid photovoltaics, based on typical daily photovoltaic and load conditions of the regional power grid.

[0097] Figure 5 This invention proposes an equivalent net load of the power system after unit regulation in a scheduling method for a regional power grid photovoltaic power system based on a thermal-solar-storage system.

[0098] Figure 6 This is a simulation diagram of scheduling scenario 1 in the scheduling method for a regional power grid photovoltaic power generation system based on thermal-solar-storage systems proposed in this invention;

[0099] Figure 7 The simulation diagram shows the scheduling scenario 2 in the scheduling method of the thermal-solar-storage system for regional power grid photovoltaic proposed in this invention;

[0100] Figure 8 The simulation diagram is shown for scheduling scenario 3 in the scheduling method of thermal-solar-storage system for regional power grid photovoltaic proposed in this invention;

[0101] Figure 9 The simulation diagram shows scheduling scenario 4 in the scheduling method for a regional power grid photovoltaic system proposed in this invention. Detailed Implementation

[0102] Reference Figures 1 to 9 .

[0103] The embodiments further illustrate the scheduling method of a thermal-solar-storage system for regional power grid photovoltaic proposed in this invention.

[0104] A method for scheduling a thermal-solar-storage system for regional power grid photovoltaic systems includes the following steps:

[0105] Collect photovoltaic power forecast and load data of regional power grid, determine the configuration parameters of power system, and obtain the start-up characteristics of the main equipment of thermal power plants and units implementing gas-steam combined cycle power generation during the start-up, shutdown and peak shaving phases.

[0106] Establish models of the main equipment in the power system, including a thermal power unit output power model and an energy storage power station model.

[0107] A hierarchical scheduling model for fire storage and peak shaving is established, including an upper-level model and a lower-level model. The objective functions of the upper-level model and the lower-level model are determined, and the constraints of the upper-level model and the lower-level model are set.

[0108] The solution process of the hierarchical scheduling model for thermal power generation and energy storage peak shaving is determined. The upper-level model aims to minimize the equivalent net load fluctuation of the power system and solve for the optimal scheduling during the start-up and shutdown peak shaving phase of the units. The lower-level model aims to minimize the operating cost within the power system scheduling cycle and solve for the output scheduling of each unit. Functions that are inconvenient to be substituted into the solution of the hierarchical scheduling model for thermal power generation and energy storage peak shaving are linearized.

[0109] The obtained scheduling scheme is implemented in the regional power grid and simulated. The system's operating status is monitored in real time, and adjustments and optimizations are made according to the actual situation.

[0110] Collect photovoltaic power forecasts and load data for the regional power grid, determine power system configuration parameters, and obtain the startup characteristics of major equipment in thermal power plants and units implementing gas-steam combined cycle power generation during start-up, shutdown, and peak shaving phases, including:

[0111] Record typical daily photovoltaic and load conditions of the regional power grid;

[0112] The power system configuration parameters include 220MW of photovoltaic capacity, 44MW / 88MWh of energy storage power station capacity, and a system spinning reserve capacity of 10% of the load.

[0113] The thermal power plant is mainly equipped with three Type I gas turbines and one Type II gas turbine. The Type I gas turbines are 6B type gas turbines, numbered 1, 2, and 3 respectively. The 6B type units are old models with low output and poor economy. The minimum gas turbine load to meet emission requirements is controlled at 80% of the rated load, and the spinning reserve capacity is limited. Therefore, it is necessary to participate in system peak shaving through rapid start-stop when necessary. The Type II gas turbine is a 6F type gas turbine, numbered 4. The minimum load of the 6F type unit can be reduced to 30% of the rated load, so it implements the conventional spinning peak shaving mode.

[0114] Thermal power plants implement gas-steam combined cycle power generation. For units in the start-up, shutdown, and peak-shaving phases, the start-up characteristics of gas-steam combined cycle power generation in actual operation can be divided into three processes: gas turbine start-up, grid connection and power generation to reach maximum load; waste heat boiler start-up and pipe warm-up; steam turbine start-up and grid connection and power generation to reach maximum load; when the unit is shut down, it is shut down in the order of steam turbine-waste heat boiler-gas turbine. The start-up characteristics of the unit are related to the residual temperature of the unit, personnel operation, and ambient temperature.

[0115] Establish models of the main equipment in the power system, including a thermal power unit output power model and an energy storage power station model.

[0116] Thermal power unit output power model:

[0117] For the number isj The units, during the regular peak-shaving phase i Effort output at any given moment can be expressed as:

[0118] (1)

[0119] In the formula: p gt,j,i —— Gas turbine power; r i —Combined cycle steam-fuel ratio;

[0120] Based on actual operating data of thermal power plants, the output during the start-up and shutdown phases of the units can be represented by the following piecewise functions (2) and (3):

[0121] (2)

[0122] (3)

[0123] In the formula, t —The time elapsed during the unit start-up and shutdown phases; P gt,j,max —The rated maximum power of the gas turbine; P gen,j,max —The maximum power unit of the combined cycle unit, ; , —Climb rate during gas turbine startup and shutdown; —Maximum permissible output power; , —Climb rate during turbine startup and shutdown; i a , i b , i c , i d —Time points during the unit startup phase i e , i f , i g —Shutdown points for steam turbines, waste heat boilers, and gas turbines;

[0124] Energy storage power station model:

[0125] Energy storage power stations i The state of charge at time t is shown in the following equation:

[0126] (4)

[0127] In the formula,S i —— i The state of charge of the energy storage power station at all times; —Data collection interval; E bn —Rated capacity of the energy storage power station; P bat,cha,i-1 , P bat,dis,i-1 —The charging and discharging power of the energy storage station at the previous moment; h bat,cha , h bat,dis — The charging and discharging efficiency of energy storage power stations.

[0128] A hierarchical scheduling model for peak shaving based on thermal power, solar power, and energy storage is established, comprising an upper-level model and a lower-level model. The objective functions of both the upper and lower-level models are determined, and their constraints are set, including:

[0129] Based on photovoltaic day-ahead power forecasting and the dynamic model of start-up and shutdown peak-shaving units, an upper-level optimization model is established to optimize the downtime and downtime duration of start-up and shutdown peak-shaving units with the objective of minimizing the standard deviation of the equivalent net load fluctuation after adjustment by the start-up and shutdown peak-shaving units; for units numbered as j , x The system is constantly shut down, and the downtime is [duration]. y The power output sequence of the generating units can be temporarily represented as:

[0130] (5)

[0131] In the formula, i down , i down,end , i up , i up,end —The time points of the unit shutdown and startup process are shown in equation (6):

[0132] (6)

[0133] Objective function of the upper-level model:

[0134] (7)

[0135] In the formula, N T — Scheduling cycle; P load,i —— i System load at all times —The adjusted load of the system at time i; P PV,max,i—— i Predicted photovoltaic power at any given time; P gen,j,i —— i Start and stop peak-shaving units at any time contribution;

[0136] The constraints of the upper-level model are as follows:

[0137] Considering that the unit needs to complete the restart operation within the remaining time of the scheduling cycle after shutdown, the constraints on the unit's shutdown time and duration are as follows: (8)

[0138] A multi-energy complementary optimal scheduling model based on thermal, solar, and storage energy is established in the lower-level model. The optimization objective is to achieve the optimal economic operation of the power system within the scheduling cycle. The objective function is: (9)

[0139] In the formula, C gen —The cost of generating electricity from thermal power plants; C bat —Operating costs of energy storage power stations; C PV —Penalties for curtailment of solar power; C net —The cost of purchasing electricity from external networks;

[0140] The power generation cost of a thermal power plant mainly considers the unit's fuel cost and start-up and shutdown costs, as shown in equations (10) and (11) below; for the first type of unit, since its power regulation range is very small, the fuel cost can be simplified to a linear function of power:

[0141] (10)

[0142] (11)

[0143] In the formula, N gen —Number of generating units; C gas,j,i —Fuel costs of the generating unit; d gen —Power generation cost of Type I generating units; a, b, c —Fuel consumption coefficient; d gas —Local natural gas price; n ss —Number of unit start-ups and shutdowns; d ss —Cost of a single start-up and shutdown of the unit;

[0144] The operating cost of an energy storage power station is:

[0145] (12)

[0146] In the formula, d bat —The charging and discharging cost coefficient of energy storage;

[0147] The penalty for curtailment of solar power is:

[0148] (13)

[0149] In the formula, P PV,loss,i —— i The amount of solar power curtailed at any given time; d PV —Levelized cost of electricity (LCOE) over the entire lifecycle of a photovoltaic power plant;

[0150] Cost of purchasing electricity from external networks:

[0151] (14)

[0152] In the formula, P net,i —— i Power purchased from the external network at any given time; d net , i —— i The State Grid electricity price is based on time-of-use pricing.

[0153] The constraints of the lower-level model include:

[0154] Neglecting system network losses, the sum of the power output from thermal power, solar power, and energy storage, and the purchased power, equals the real-time load; power system power balance constraints: (15)

[0155] In the formula, P net,i —— i Electricity purchase via the national online shopping platform; P pv,loss,i for i Real-time solar curtailment volume;

[0156] For units in the start-up and shutdown peak shaving phase, their output power is consistent with equation (5); for units in the rotating peak shaving phase, their output power is limited by the unit's ramp rate and the upper limit of the rated maximum output. This represents the maximum permissible deviation of the j-th thermal power unit. This represents the minimum permissible deviation of the j-th thermal power unit; dynamic constraints of thermal power plant units:

[0157] (16)

[0158] This indicates the total number of thermal power units and the standby capacity constraints of thermal power plants.

[0159] (17)

[0160] In the formula, U gen,j,i --unit j exist i The state variable at that time, 0 represents the unit being shut down, and 1 represents the unit being running; Ri --system i Backup capacity requirements at times;

[0161] Energy storage power station charging and discharging power constraints:

[0162] (18)

[0163] In the formula, P bat,cha,max 、P bat,dis,max —Maximum charging and discharging power of the energy storage power station, in MW; U cha,i , U dis,i —A Boolean variable indicating the charging and discharging of an energy storage power station, where 0 indicates that it is not in a charging or discharging state, and 1 indicates that it is in a charging or discharging state;

[0164] State of charge constraints for energy storage power stations:

[0165] (19)

[0166] In the formula, S max 、S min —The upper and lower limits of the state of charge of energy storage power stations; S 1 , S NT —The state of charge at the beginning and end of the scheduling cycle;

[0167] External electricity purchase restrictions:

[0168] (20)

[0169] In the formula, P net,max —The upper limit of the power capacity for purchasing electricity from external networks.

[0170] The solution process for the hierarchical scheduling model for thermal power generation and energy storage peak shaving is determined. The upper-level model aims to minimize the equivalent net load fluctuation of the power system to solve for the optimal scheduling during the start-up and shutdown peak shaving phases of the units. The lower-level model aims to minimize the operating cost within the power system scheduling cycle to solve for the output scheduling of each unit. Functions that are inconvenient to be incorporated into the solution of the hierarchical scheduling model for thermal power generation and energy storage peak shaving are linearized, including:

[0171] The upper-level model first obtains the net load characteristics through the photovoltaic day-ahead power prediction model, which serves as the input to the upper-level model; a combination algorithm of FCM and LSTM is used to achieve short-term prediction of photovoltaic power; the Long Short-Term Memory Neural Network (LSTM) has a special memory and forgetting mode to flexibly adapt to the temporal characteristics of the learning task; Fuzzy C-means clustering (FCM) can classify the weather type of the samples according to the fluctuation level of the total horizontal irradiance of historical samples, thus making it convenient to select similar days as training samples for LSTM.

[0172] Based on the number of start-up and shutdown peak-shaving units, the particle swarm optimization algorithm is used to minimize the standard deviation of the equivalent net load fluctuation of the power system after unit regulation. The shutdown time and duration of each start-up and shutdown unit are solved iteratively. The equivalent net load after regulation by the previous start-up and shutdown peak-shaving unit is used as the input for the next cycle. With the goal of optimizing the operating economy within the power system scheduling cycle, the output scheduling of each unit is solved by calling the CPLEX solver through Matlab.

[0173] The following functions that are inconvenient to be incorporated into the solution of the fire-storage peak-shaving hierarchical scheduling model are linearized, including:

[0174] Unit fuel costs:

[0175] First of all C gas,j,i Convert to n segmental linear functions f (P gt,j,i ), the dividing points are Introducing variables w k,j,i Will P gt,j,i ,and f ( P gt,j,i ) is represented by equation (21). —The state identifier variable of the j-th thermal power unit in the k-th segment under the i-th operating state. —The state identifier variable of the j-th thermal power unit in the (r-1)-th segment under the i-th operating state. —The state identifier variable of the j-th thermal power unit in the first segment under the i-th operating state. —The state identifier variable of the j-th thermal power unit in the r-th segment under the i-th operating state. —The two-state identifier variable of the j-th thermal power unit in the n-th segment under the i-th operating state. This represents the output power of the i-th unit under the j-th type of equipment during the first scheduling period. This represents the output power of the i-th unit under the j-th type of equipment during the r-th scheduling period. Let represent the output power of the i-th unit under the j-th type of equipment during the (n+1)-th scheduling period. The constraints of each variable are shown in equation (22):

[0176] (twenty one)

[0177] (twenty two)

[0178] Energy storage power station charging and discharging power constraints:

[0179] The constraints are transformed using the Big-M method:

[0180] (twenty three)

[0181] In the formula, M is a sufficiently large constant.

[0182] The obtained scheduling scheme is implemented in the regional power grid and simulated. The operating status of the power system is monitored in real time, and adjustments and optimizations are made according to the actual situation, including:

[0183] Based on the current photovoltaic forecast and peak-shaving unit start-up and shutdown model, the particle swarm optimization algorithm is used to solve the operation scheduling of start-up and shutdown units of thermal power plants according to the number of start-up and shutdown peak-shaving units put into operation. When multiple start-up and shutdown units are put into operation, the order in which units of different capacities participate in the cyclic optimization has an impact on the optimization results. Since the three units equipped in the power plant have similar installed capacities, this impact can be ignored and therefore they are not distinguished. After putting into operation different numbers of start-up and shutdown peak-shaving units, the equivalent net load result of the power system is solved.

[0184] Based on the number of peak-shaving units put into operation, the upper-level model can generate three optimal scheduling schemes. Based on these different scheduling schemes, four scheduling scenarios are formed, including scheduling scenario 1. ~ 3 corresponds to 3 scheduling schemes respectively. Scheduling scenario 4 corresponds to the scenario where no peak-shaving units are started or stopped. Simulations are performed on these 4 scheduling scenarios to obtain the operating cost composition of the power system under each scheduling scenario.

[0185] Reference Figure 2 :

[0186] Solution process for the hierarchical scheduling model of thermal power generation and energy storage for peak shaving: Total number of peak shaving units to be started and stopped n ssA combined algorithm of FCM and LSTM is used to achieve short-term prediction of photovoltaic power, obtaining the predicted photovoltaic power; the number of peak-shaving units put into regulation is set to... i initial value i =0, iteratively solve for the downtime and downtime duration of the start-up and shutdown units. x,y When solving in each iteration... i Add one more, until... i ≤ n ss If the condition is not met, terminate the loop solution; based on the number of peak-shaving units started and stopped for regulation. i Based on the differences, obtain the equivalent net load curve of the power system, and sequentially increase the number of start-up and shutdown peak-shaving units put into regulation. i When the number of peak-shaving units put into regulation i ≤Number of peak-shaving units started and stopped n ss The algorithm iteratively solves the objective function using a particle swarm optimization (PSO) approach, aiming to minimize the standard deviation of the equivalent net load fluctuation of the power system after regulation by the start-up and shutdown of peak-shaving units. This process is then applied to the number of start-up and shutdown peak-shaving units involved in the regulation. i To find the optimal solution, the equivalent net load of the power system after regulation by the previous start-up and shutdown of the peak-shaving unit is used as the input for the next cycle. The downtime and downtime duration of each start-up and shutdown unit are then iteratively calculated. x,y And the output of each start-up and shutdown peak-shaving unit during the start-up and shutdown peak-shaving phases; when the number of start-up and shutdown peak-shaving units put into regulation. i ≥ Number of peak-shaving units started and stopped n ss When the loop is terminated, the output of each start-stop peak-shaving unit during the start-stop peak-shaving phase is obtained from the solution. The objective function is to minimize the operating cost within the power system dispatch cycle. The output of each start-stop peak-shaving unit during the normal peak-shaving phase and the output of each subsystem are solved by calling the CPLEX solver through Matlab.

[0187] Reference Figure 5 :

[0188] Due to the large proportion of photovoltaic installations, the peak-to-valley difference of the power system's net load is too high. The more start-stop peak-shaving units are put into operation, the smoother the equivalent net load of the power system after regulation, and the smaller the peak-to-valley difference of the power system. However, after putting into operation 3 start-stop peak-shaving units, the overall net load level of the power system is relatively low, basically between 0 and 50MW, thus compressing the output space of other power sources in the power system. Based on the number of start-stop peak-shaving units put into operation, the upper-level model can generate 3 optimal start-stop scheduling schemes. The standard deviation of the original net load of the power system before regulation is 76.56MW. After regulation by putting into operation 3 start-stop peak-shaving units, the standard deviation of the net load of the power system is reduced to a minimum of 25.30MW.

[0189] The scheduling results of starting and stopping peak-shaving units are shown in Table 1 below:

[0190]

[0191] Reference Figures 6-9 :

[0192] Analysis of the simulation diagrams for the four dispatching scenarios reveals that, on the one hand, during periods without photovoltaic (PV) power generation, the power system primarily supplies regional loads through thermal power plants. During peak PV power generation periods, the power system reduces the minimum output of thermal power plants by shutting down some Type I units, and enhances the PV absorption capacity of the power system in conjunction with energy storage devices. Simultaneously, Type II units are kept operational to meet the power system's reserve spinning capacity requirements and ensure power supply reliability. On the other hand, during off-peak hours, the power system purchases electricity from the external grid; during peak hours, energy storage stations discharge to reduce the power system's external electricity purchases.

[0193] In dispatch scenario 2, during peak photovoltaic power generation periods, thermal power plants only maintain one unit, resulting in a low output level. This forces the power system to purchase electricity from the external grid to balance the load gap caused by photovoltaic fluctuations. In dispatch scenario 3, thermal power plants put three start-stop peak-shaving units into operation, leading to excessive output from thermal power plants during periods without photovoltaic power generation, which reduces the power system's ability to purchase off-peak electricity.

[0194] The composition of power system operating costs under various dispatching scenarios is shown in Table 2 below:

[0195]

[0196] Compared to dispatch scenario 4, dispatch scenario 1, after deploying one start-stop peak-shaving unit, although increased the start-stop costs of thermal power plants, reduced their power generation, and squeezed their profits, resulted in a 11.80% decrease in the power system's curtailment rate, leading to a 14.14% reduction in the overall operating cost of the power system. Dispatch scenario 2, based on dispatch scenario 1, added one more start-stop peak-shaving unit, further reducing the power system's curtailment rate and operating cost by 13.15% and 9.71%, respectively. In contrast, dispatch scenario 3, with its excessive deployment of start-stop peak-shaving units, resulted in a less favorable performance in both the power system's curtailment rate and operating cost compared to dispatch scenario 2.

[0197] Compared with the single rotating peak shaving mode, the curtailment rate and operating cost of the power system under the hierarchical scheduling of thermal power and energy storage peak shaving decreased by 34.95% and 22.48%, respectively. The results show that the proposed hierarchical scheduling model of thermal power and energy storage peak shaving can effectively improve the grid's absorption of photovoltaic power in the region and reduce the operating cost of the power system.

[0198] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for scheduling a firelight storage system for a regional grid photovoltaic, characterized in that, The method comprises the following steps: Collecting photovoltaic predicted power and load data of a regional power grid, determining configuration parameters of a power system, obtaining main equipment of a thermal power plant and start-up characteristics of a unit implementing gas-steam combined cycle power generation during a start-stop peak regulation stage; Establishing a power system main equipment model, wherein the power system main equipment model comprises a thermal power unit output power model and an energy storage power station model; The thermal power unit output power model: For the unit numbered j , the output at the time of the conventional peak regulation phase can be expressed as: i ​ (1) In the formula: p gt,j,i —— Gas turbine power; r i Combined cycle steam to gas power ratio; In combination with actual operation data of the thermal power plant, output of the unit during the start-stop stage can be respectively represented by the following segmented functions (2) and (3): （2） （3） In the formula, t - the time elapsed during the start-up and shut-down phases of the unit; P gt,j,max - the rated maximum power of the combustion engine; P gen,j,max - maximum power of combined cycle unit, ; , - ramp rate of gas turbine during start-up and shut-down; , - ramp rate of steam turbine during start-up and shut-down; i a , i b , i c , i d - time nodes during start-up of unit, i e , i f , i g - shut-down nodes of steam turbine, waste heat boiler and gas turbine; The energy storage power station model: The energy storage power station i The state of charge at the moment is shown in the following formula: （4） In the formula, S i — i State of charge of the energy storage plant at the moment; — Data acquisition interval; E bn Rated capacity of the energy storage plant; P bat,cha,i-1 , P bat,dis,i-1 Charging and discharging power of the energy storage plant at the previous moment; h bat,cha , h bat,dis Charging and discharging efficiency of the energy storage plant; The establishment of the thermal storage peak regulation hierarchical scheduling model comprises an upper model and a lower model, the objective functions of the upper model and the lower model are determined, and the constraint conditions of the upper model and the lower model are set; The solving process of the thermal storage peak regulation hierarchical scheduling model is determined, the upper model is solved with the minimum equivalent net load fluctuation of the power system as the target to obtain optimal scheduling of the unit during the start-stop peak regulation stage, the lower model is solved with the minimum operation cost of each unit during the power system scheduling period as the target to obtain output scheduling of each unit, and functions that are inconvenient to be brought into the thermal storage peak regulation hierarchical scheduling model are linearly processed; The scheduling scheme obtained by solving is implemented into the regional power grid and simulation test is performed, the operation state of the system is monitored in real time, and adjustment and optimization are performed according to actual conditions. 2.The firelight storage system scheduling method for regional grid-oriented photovoltaic according to claim 1, wherein, Collecting photovoltaic predicted power and load data of a regional power grid, determining configuration parameters of a power system, obtaining main equipment of a thermal power plant and start-up characteristics of a unit implementing gas-steam combined cycle power generation during a start-stop peak regulation stage, comprising: Recording typical daily photovoltaic and load conditions of the regional power grid; The configuration parameters of the power system comprise 220 MW of photovoltaic installation, 44 MW / 88 MWh of energy storage power station installation, and 10% of the load of system rotating reserve capacity; The main equipment of the thermal power plant is equipped with three first-type gas turbines and one second-type gas turbine; The thermal power plant implements gas-steam combined cycle power generation, and for the unit during the start-stop peak regulation stage, the start-up characteristics of the gas-steam combined cycle power generation during actual operation process comprises three processes: the gas turbine starts, is connected to the grid and reaches maximum load; the waste heat boiler starts to warm the pipe; the steam turbine starts, is connected to the grid and reaches maximum load; when the unit is stopped, the steam turbine, the waste heat boiler and the gas turbine are sequentially stopped in this order.

3. The method of claim 2, wherein, The establishment of the thermal storage peak regulation hierarchical scheduling model comprises an upper model and a lower model, the objective functions of the upper model and the lower model are determined, and the constraint conditions of the upper model and the lower model are set, comprising: Based on photovoltaic day-ahead power forecasting and the dynamic model of start-up and shutdown peak-shaving units, an upper-level optimization model is established to optimize the downtime and downtime duration of start-up and shutdown peak-shaving units with the objective of minimizing the standard deviation of the equivalent net load fluctuation after adjustment by the start-up and shutdown peak-shaving units; for units numbered as j , x The system is constantly shut down, and the downtime is [duration]. y The power output sequence of the generating units can be temporarily represented as: （5） In the formula, i down , i down,end , i up , i up,end The time nodes of the unit shutdown and startup process are shown in formula (6): （6） The objective function of the upper model: （7） In the formula, N T — Scheduling cycle; P load,i —— i System load at all times; P PV,max,i —— i Predicted photovoltaic power at any given time; P gen,j,i —— i Start and stop peak-shaving units at any time of effort; The constraint conditions of the upper model: Considering that the unit needs to complete the re-start action within the remaining time of the scheduling period after being stopped, the constraint of the stop time and the stop duration of the unit is: （8） In the lower model, a thermal photovoltaic storage multi-energy complementary optimization scheduling model is established, the power system operation economy within the scheduling period is optimized as the target, and the objective function is: （9） wherein C gen — cost of electricity generation in a thermal power plant; C bat — cost of operation of an energy storage plant; C PV — penalty for curtailment of photovoltaic power; C net — cost of electricity purchase from the external grid; The power generation cost of thermal power plants mainly considers the unit fuel cost and the start-stop cost, as shown in the following formulas (10) and (11); for the first type of unit, since its power regulation range is small, the fuel cost can be simplified as a linear function of power: （10） （11） In the formula, N gen — the number of units; C gas,j,i — the fuel cost of the unit; d gen — the power generation cost of the 6B type unit; a, b, c — the fuel consumption coefficient; d gas — the local natural gas selling price; n ss — the number of unit start-stop times; d ss — the unit single start-stop cost; The operation cost of the energy storage power station is: （12） In the formula, d bat — the charging and discharging cost coefficient of energy storage; The photovoltaic light abandonment penalty is: （13） In the formula, P PV,loss,i — i The amount of abandoned light of photovoltaic at the moment; d PV — The degree of electric cost of photovoltaic power station in the whole life cycle The external grid purchase cost is: （14） In the formula, P net,i — i The external network power at the moment; d net , i — i The national network price at the moment, adopts time-of-use electricity price; The constraint conditions of the lower layer model include: When the system network loss is not considered, the sum of the power output of the thermal, light and storage units and the purchase power is equal to the real-time load; the power balance constraint of the power system: （15） In the formula, P net,i — i The instantaneous state grid electricity purchase amount; P pv,loss,i For i The instantaneous state photovoltaic light waste amount; For the units in the start-stop peak regulation stage, the output power is consistent with formula (5); for the units in the rotating peak regulation stage, the output power is limited by the climbing rate and the upper limit of the rated maximum output; the thermal power plant dynamics constraint is: （16） The thermal power plant reserve capacity constraint is: （17） In the formula, U gen,j,i --unit j exist i The state variable at that time, 0 represents the unit being shut down, and 1 represents the unit being running; Ri --system i Backup capacity requirements at times; The charge and discharge power constraint of the energy storage power station is: （18） In the formula, P bat,cha,max 、P bat,dis,max - Maximum charge and discharge power of the energy storage power station, MW; U cha,i , U dis,i - Charge and discharge identification Boolean variable of the energy storage power station, 0 represents not in the state of charging and discharging, and 1 represents in the state of charging and discharging. The state of charge constraint of the energy storage power station is: （19） wherein S max 、S min - upper and lower limits of the state of charge of the energy storage plant; S 1 , S NT - state of charge at the beginning and end of the dispatching period; The external grid purchase constraint is: （20） In the formula, P net,max — upper limit of the external grid purchase power.

4. The method of claim 3, wherein, The solving process of the thermal storage peak regulation hierarchical scheduling model is determined, the upper layer model solves the optimal scheduling of the start-stop peak regulation stage of the unit with the minimum equivalent net load fluctuation of the power system as the target; in the lower layer model, the output scheduling of each unit is solved with the minimum operation cost in the power system scheduling period as the target; the functions that are not convenient to bring into the thermal storage peak regulation hierarchical scheduling model for solving are linearized, including: The upper layer model first obtains the net load characteristics through the photovoltaic day-ahead power prediction model as the input of the upper layer model; the combination algorithm of FCM and LSTM is used to realize the short-term prediction of photovoltaic power; According to the number of units put into start-stop peak regulation, the particle swarm algorithm is used to minimize the standard deviation of the equivalent net load fluctuation of the power system after unit regulation, and the shutdown time and shutdown duration of each start-stop unit are solved, the equivalent net load after the adjustment of the last start-stop peak regulation unit is taken as the input of the next cycle, and the output scheduling of each unit is solved by calling the CPLEX solver in Matlab with the optimal operation economy of the power system in the scheduling period as the target; The following functions that are not convenient to bring into the thermal storage peak regulation hierarchical scheduling model for solving are linearized, including: The unit fuel cost is: First, we have C gas,j,i converted to n piecewise linear functions f (P gt,j,i ), with breakpoints at , and introducing variables w k,j,i We have P gt,j,i , and f ( P gt,j,i ) expressed as in equation (21), with variable constraints as in equation (22): （21） （22） The charge and discharge power constraint of the energy storage power station is: The constraint is converted to: （23） In the formula, M is a large enough constant.

5. The method of claim 4, wherein, The solved scheduling scheme is implemented into the regional power grid and simulation test is carried out, the operation state of the power system is monitored in real time, and adjustment and optimization are carried out according to the actual situation, including: Based on the day-ahead photovoltaic prediction and peak regulation unit start-stop model, the particle swarm algorithm is used to solve the operation scheduling of the thermal power plant start-stop unit according to the number of units put into start-stop peak regulation; after putting different number of start-stop peak regulation units, the equivalent net load result of the power system is solved; According to the number of units put into start-stop peak regulation, the upper layer model can generate three optimal scheduling schemes, according to different scheduling schemes, four scheduling scenarios are formed, and the operation cost composition of the power system under various scheduling scenarios is obtained by simulating the four scheduling scenarios.

Images