Regional power grid photovoltaic oriented flame light storage system scheduling method

Abstract

The invention discloses a regional power grid photovoltaic oriented flame-light storage system scheduling method, which relates to the technical field of power system scheduling, and is characterized by comprising the steps of data acquisition and prediction, establishment of a main equipment model of a power system, establishment of a flame-light storage peak regulation hierarchical scheduling model comprising an upper layer model and a lower layer model, and determination of a target function. Constraint conditions are set, the solving process of the fire storage peak regulation hierarchical scheduling model is determined, the upper-layer model solves the optimal scheduling of the unit start-stop peak regulation stage with the minimum equivalent net load fluctuation of the power system as the target, and the lower-layer model solves the output scheduling of each unit with the minimum operation cost in the scheduling period of the power system as the target. Performing linearization processing on a function which is inconvenient to be brought into the fire storage peak regulation hierarchical scheduling model for solving, and finally performing a simulation test by taking a real scene as a case; the method has the effects of improving the photovoltaic absorption level of the regional power grid and reducing the operation cost of a power system.

Description

Technical Field

[0001] The present invention relates to the technical field of power system dispatching, and more specifically, to a dispatching method for a photovoltaic thermal energy storage system for a regional power grid. Background Art

[0002] With the rapid development of renewable energy, the proportion of photovoltaic power generation in the power grid continues to increase, but the intermittent and uncertain output of photovoltaic power generation has brought great challenges 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 there are environmental emission problems. The introduction of thermal photovoltaic storage systems provides a new technical means to balance the fluctuations in photovoltaic output. Therefore, studying a scheduling method for thermal photovoltaic storage systems for regional power grid photovoltaics is of great significance to improving the overall operation efficiency and stability of the power grid. Summary of the invention

[0003] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a method for scheduling a photovoltaic thermal energy storage system for a regional power grid.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A method for dispatching a photovoltaic power storage system for a regional power grid includes the following steps:

[0006] Collect the photovoltaic power forecast and load data of the regional power grid, determine the configuration parameters of the power system, and obtain the startup characteristics of the main equipment of the thermal power plant and the units implementing gas-steam combined cycle power generation during the start-stop peak regulation stage;

[0007] Establishing a main equipment model of the power system, wherein the main equipment model of the power system includes a thermal power unit output power model and an energy storage power station model;

[0008] Establish a hierarchical dispatching model for peak load regulation of thermal storage, including an upper model and a lower model, determine the objective functions of the upper model and the lower model, and set the constraints of the upper model and the lower model;

[0009] Determine the solution process of the thermal storage peak-shaving hierarchical scheduling model. The upper model solves the optimal scheduling of the start-stop peak-shaving stage of the unit with the goal of minimizing the equivalent net load fluctuation of the power system; the lower model solves the output scheduling of each unit with the goal of minimizing the operating cost within the power system scheduling cycle; linearize the functions that are inconvenient to be brought into the thermal storage peak-shaving hierarchical scheduling model;

[0010] The solved dispatching scheme is implemented in the regional power grid and a simulation test is carried out to monitor the operating status of the system in real time, and to make adjustments and optimizations based on actual conditions.

[0011] Preferably, the photovoltaic predicted power and load data of the regional power grid are collected, the configuration parameters of the power system are determined, and the startup characteristics of the main equipment of the thermal power plant and the units implementing the gas-steam combined cycle power generation during the start-stop peak regulation stage are obtained, including:

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

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

[0014] The main equipment of the thermal power plant is equipped with three first-type gas turbines and one second-type gas turbine;

[0015] Thermal power plants implement gas-steam combined cycle power generation. For units in the start-stop and peak-shaving stage, the startup characteristics of gas-steam combined cycle power generation during actual operation are divided into three processes: gas turbine startup, grid-connected power generation and reaching maximum load; waste heat boiler startup and pipe warming; steam turbine startup and run-up, grid-connected power generation and reaching 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 main equipment model of the power system is established, wherein the main equipment model of the power system includes an output power model of a thermal power unit and an energy storage power station model;

[0017] Output power model of thermal power unit:

[0018] For the unit numbered j, its output at time i during the conventional peak load regulation phase can be expressed as:

[0019] P gen,j,i =P gt,j,i (1+r i ) (1)

[0020] Where: p gt,j,i ——gas turbine power; r i ——Combined cycle steam-combustion power ratio;

[0021] Combined with the actual operation data of thermal power plants, the output of the unit during the start-up and shutdown stages can be expressed by the following piecewise functions (2) and (3) respectively:

[0022]

[0023] Where, t is the time between the start and stop phases of the unit; P gt,j,max ——Rated maximum power of the engine; P gen,j,max ——Maximum power of combined cycle unit, P gen,j,max =P gt,j,max +P wt,j,max; ——Ramp rate when starting and stopping the gas turbine; ——Ramp rate when the turbine is started and shut down; i a 、i b 、i c 、i d ——Time nodes during the unit startup phase, i e 、i f 、i g ——Shutdown nodes of steam turbines, waste heat boilers and gas turbines;

[0024] Energy storage power station model:

[0025] The charge state of the energy storage power station at time i is as follows:

[0026]

[0027] In the formula, S i ——The charge state of the energy storage power station at time i; Δt——data collection interval; E bn ——Rated capacity of energy storage power station; P bat,cha,i-1 , P bat,dis,i-1 ——The charging and discharging power of the energy storage power station at the previous moment; η bat,cha , η bat,dis ——Charging and discharging efficiency of energy storage power stations.

[0028] Preferably, establishing a thermal-solar-storage peak-shaving hierarchical scheduling model includes an upper model and a lower model, determining the objective functions of the upper model and the lower model, and setting constraints of the upper model and the lower model, including:

[0029] Based on the photovoltaic day-ahead power forecast and the dynamic model of the start-stop peak-shaving unit, an upper-level optimization model is established to optimize the downtime and downtime of the start-stop peak-shaving unit with the goal of minimizing the standard deviation of the equivalent net load fluctuation after the start-stop peak-shaving unit is adjusted; for the unit numbered j, shut down at time x, and with a downtime of y, its output sequence can be temporarily expressed as:

[0030]

[0031] In the formula, i down 、i down,end 、i up 、i up,en ——The time nodes of the unit shutdown and startup process, as shown in formula (6):

[0032]

[0033] The objective function of the upper model is:

[0034]

[0035] Where N T ——Scheduling period; P load,i ——system load at time i; P PV,max,i ——PV power prediction value at time i; P gen,j,i ——The output of peak load regulating unit j started and stopped at time i;

[0036] The constraints of the upper model are:

[0037] Considering that the unit needs to complete the restart action within the remaining time of the scheduling cycle after shutdown, the constraints on the shutdown time and shutdown duration of the unit are:

[0038]

[0039] In the lower model, a multi-energy complementary optimization dispatching model of thermal, solar and storage is established, and the optimization is carried out with the goal of optimizing the economic efficiency of the power system operation within the dispatching period. The objective function is:

[0040] f COST =C gen +C bat +C PV +C net (9)

[0041] In the formula, C gen ——The cost of electricity generation from thermal power plants; C bat ——Operation cost of energy storage power station; C PV ——Penalty for photovoltaic abandonment; C net ——The cost of purchasing electricity from the external grid;

[0042] The power generation cost of a thermal power plant mainly considers the unit fuel cost and start-up and shutdown cost, as shown in the following equations (10) and (11). For the first type of unit, due to its small power regulation range, the fuel cost can be simplified as a linear function of power:

[0043]

[0044] Where N gen ——Number of units; C gas,j,i ——Unit fuel cost; δ gen ——Generation cost of the first type of unit; a, b, c——fuel consumption coefficient; δ gas ——Local natural gas selling price; n ss ——Number of unit starts and stops; δ ss ——Single start-up and shutdown cost of the unit;

[0045] The operating cost of the energy storage power station is:

[0046] Cbat =(P bat,cha,i +P bat,dis,i )Δtδ bat (12)

[0047] In the formula, δ bat ——Charging and discharging cost coefficient of energy storage;

[0048] The penalty for photovoltaic abandonment is:

[0049]

[0050] Where P PV,loss,i ——the amount of photovoltaic power abandoned at time i; δ PV ——The cost of electricity per kilowatt-hour during the entire life cycle of a photovoltaic power station;

[0051] Cost of purchasing electricity from external network:

[0052]

[0053] Where P net,i ——The power purchased from the external grid at time i; δ net,i ——The electricity price of the State Grid at time i, using the time-of-use electricity price;

[0054] The constraints of the lower model include:

[0055] When the system network loss is ignored, the sum of the thermal, solar, and storage output and purchased electricity of the power system is equal to the real-time load; the power balance constraint of the power system is:

[0056]

[0057] Where P net,i ——The amount of electricity purchased online at time i; P pv,loss,i is the amount of photovoltaic abandoned light at time i;

[0058] For the units in the start-stop peak-shaving stage, their output power is consistent with formula (5); for the units in the rotating peak-shaving stage, their output power is limited by the unit climbing rate and the rated maximum output upper limit; the dynamic constraints of the thermal power plant units are:

[0059]

[0060] Reserve capacity constraints of thermal power plants:

[0061]

[0062] Where U gen,j,i ——The state variable of unit j at time i, 0 represents the unit is shut down, 1 represents the unit is running; R i ——the spare capacity of the system at time i;

[0063] Energy storage power station charging and discharging power constraints:

[0064]

[0065] Where P bat,cha,max , P bat,dis,max ——The maximum charging and discharging power of the energy storage station, MW; U cha,i , U dis,i ——Boolean variable for the charge and discharge identification of the energy storage power station, 0 means it is not in the charge or discharge state, and 1 means it is in the charge or discharge state;

[0066] State of charge constraints of energy storage power stations:

[0067]

[0068] In the formula, S max , S min ——The upper and lower limits of the state of charge of the energy storage power station; S1, S NT ——Charge status at the beginning and end of the dispatch cycle;

[0069] Restrictions on purchasing electricity from the Internet:

[0070] 0≤P net,i ≤P net,max (20)

[0071] Where P net,max ——The upper limit of electricity purchased from the external grid.

[0072] Preferably, the solution process of the thermal storage peak-shaving hierarchical scheduling model is determined. The upper model solves the optimal scheduling of the start-stop peak-shaving stage of the unit with the goal of minimizing the equivalent net load fluctuation of the power system; the lower model solves the output scheduling of each unit with the goal of minimizing the operating cost within the power system scheduling cycle; the functions that are inconvenient to be brought into the thermal storage peak-shaving hierarchical scheduling model are linearized, including:

[0073] The upper model first obtains the net load characteristics through the photovoltaic day-ahead power prediction model as the input of the upper model; the combined algorithm of FCM and LSTM is used to realize the short-term prediction of photovoltaic power;

[0074] According to the number of start-stop peak-shaving units put into operation, the particle swarm algorithm is used to minimize the standard deviation of the equivalent net load fluctuation of the power system after the unit adjustment, and the downtime and downtime duration of each start-stop unit are solved cyclically. The equivalent net load adjusted by the previous start-stop peak-shaving unit is used as the input of the next cycle. With the goal of optimizing the operating economy within the power system dispatch cycle, the CPLEX solver is called through Matlab to solve the output dispatch of each unit;

[0075] The following functions that are inconvenient to be brought into the fire storage peak load stratified scheduling model for solution are linearized, including:

[0076] Unit fuel cost:

[0077] First, C gas,j,i Convert to n-segment linear function f(P gt,j,i ), the sub-point is b 1,j,i ≤ b 2,j,i ≤…≤b n,j,i ≤ b n+1,j,i , introduce the variable w k,j,i P gt,j,i , and f(P gt,j,i ) is expressed as formula (21), and the constraints of each variable are shown in formula (22):

[0078]

[0079] Energy storage power station charging and discharging power constraints:

[0080] Using the big-M method, the constraints are transformed into:

[0081]

[0082] Where M is a sufficiently large constant.

[0083] Preferably, the obtained dispatching scheme is implemented in the regional power grid and a simulation test is carried out to monitor the operation status of the power system in real time and adjust and optimize it according to the actual situation, including:

[0084] Based on the day-ahead photovoltaic forecast and the start-stop model of the peak-shaving units, the particle swarm algorithm is used to solve the operation and dispatch of the start-stop units of the thermal power plant according to the number of start-stop peak-shaving units put into use; after putting in different numbers of start-stop peak-shaving units, the equivalent net load results of the power system are solved;

[0085] According to the number of peak-shaving units put into operation, the upper-level model can generate three optimal dispatching schemes. According to different dispatching schemes, four dispatching scenarios are formed. These four dispatching scenarios are simulated to obtain the operating cost composition of the power system under various dispatching scenarios.

[0086] Compared with the prior art, the present invention has the following beneficial effects:

[0087] Improved the stability of the regional power grid. Through the reasonable scheduling of the thermal photovoltaic storage system, it can effectively balance the intermittency and volatility of photovoltaic power generation and ensure the stability of the voltage and frequency of the power grid.

[0088] Improved the absorption capacity of renewable energy, giving priority to using photovoltaic power generation to meet load demand, and storing excess electricity through energy storage systems, reducing the phenomenon of abandoned light;

[0089] Reduced operating costs, by optimizing scheduling strategies and rationally allocating the output of thermal power generation and energy storage systems, reducing fuel consumption and equipment wear;

[0090] It has good adaptability and scalability, and can be adjusted and optimized according to the characteristics and needs of different regional power grids. BRIEF DESCRIPTION OF THE DRAWINGS

[0091] Figure 1 The present invention proposes a flow chart of a method for dispatching a photovoltaic thermal energy storage system for a regional power grid;

[0092] Figure 2 The present invention proposes a solution flow chart of the thermal-solar-storage peak-shaving hierarchical scheduling model in a thermal-solar-storage system scheduling method for regional power grid photovoltaics

[0093] Figure 3 The present invention proposes a gas-steam combined cycle unit startup characteristic in a method for dispatching a photovoltaic thermal energy storage system for a regional power grid;

[0094] Figure 4 The present invention proposes a method for dispatching a photovoltaic power storage system for a regional power grid, which includes typical daily photovoltaic and load conditions of the regional power grid;

[0095] Figure 5 The invention proposes a method for dispatching a photovoltaic thermal energy storage system for a regional power grid, and provides an equivalent net load of the power system after the unit is adjusted;

[0096] Figure 6 The present invention proposes a simulation diagram of scheduling scenario 1 in a method for scheduling a thermal photovoltaic storage system for regional power grid photovoltaics;

[0097] Figure 7 The present invention proposes a simulation diagram of scheduling scenario 2 in a method for scheduling a thermal photovoltaic storage system for regional power grid photovoltaics;

[0098] Figure 8 The present invention proposes a simulation diagram of scheduling scenario 3 in a method for scheduling a photovoltaic thermal energy storage system for a regional power grid;

[0099] Fig. 9 The present invention proposes a simulation diagram of scheduling scenario 4 in a scheduling method for a photovoltaic thermal energy storage system for a regional power grid. DETAILED DESCRIPTION

[0100] Reference Figures 1 to 9 .

[0101] The embodiment further illustrates a method for dispatching a photovoltaic thermal energy storage system for a regional power grid proposed by the present invention.

[0102] A method for dispatching a photovoltaic power storage system for a regional power grid includes the following steps:

[0103] Collect the photovoltaic power forecast and load data of the regional power grid, determine the configuration parameters of the power system, and obtain the startup characteristics of the main equipment of the thermal power plant and the units implementing gas-steam combined cycle power generation during the start-stop peak regulation stage;

[0104] Establishing a main equipment model of the power system, wherein the main equipment model of the power system includes a thermal power unit output power model and an energy storage power station model;

[0105] Establish a hierarchical dispatching model for peak load regulation of thermal storage, including an upper model and a lower model, determine the objective functions of the upper model and the lower model, and set the constraints of the upper model and the lower model;

[0106] Determine the solution process of the thermal storage peak-shaving hierarchical scheduling model. The upper model solves the optimal scheduling of the start-stop peak-shaving stage of the unit with the goal of minimizing the equivalent net load fluctuation of the power system; the lower model solves the output scheduling of each unit with the goal of minimizing the operating cost within the power system scheduling cycle; linearize the functions that are inconvenient to be brought into the thermal storage peak-shaving hierarchical scheduling model;

[0107] The solved dispatching scheme is implemented in the regional power grid and a simulation test is carried out to monitor the operating status of the system in real time, and to make adjustments and optimizations based on actual conditions.

[0108] Collect the photovoltaic power forecast and load data of the regional power grid, determine the configuration parameters of the power system, and obtain the startup characteristics of the main equipment of the thermal power plant and the units implementing gas-steam combined cycle power generation during the start-stop peak regulation stage, including:

[0109] Record typical daily photovoltaic and load conditions of regional power grid;

[0110] The configuration parameters of the power system include 220MW of photovoltaic power generation, 44MW / 88MWh of energy storage power station, and a system spinning reserve capacity of 10% of the load;

[0111] The main equipment of the thermal power plant is equipped with 3 first-type gas turbines and 1 second-type gas turbine; the first-type gas turbine is a 6B gas turbine, which is numbered as 1, 2, and 3. The 6B unit is old, with low output and poor economy. The minimum engine load that meets emission requirements is controlled at 80% of the rated load, and the rotating reserve capacity is limited. Therefore, when necessary, it is necessary to participate in system peak regulation by rapid start and stop; the second-type gas turbine is a 6F gas turbine, which is numbered as 4. The minimum load of the 6F unit can be reduced to 30% of the rated load, so the conventional rotating peak regulation mode is implemented;

[0112] Thermal power plants implement gas-steam combined cycle power generation. For units in the start-stop and peak-shaving stage, the startup characteristics of the gas-steam combined cycle power generation during actual operation are divided into three processes: gas turbine startup, grid-connected power generation and reaching maximum load; waste heat boiler startup and pipe warming; steam turbine startup and run-up, grid-connected power generation and reaching maximum load; when the unit is shut down, it is shut down in the order of steam turbine-waste heat boiler-gas turbine; the unit's startup characteristics are related to the unit's residual temperature status, personnel operation, and ambient temperature.

[0113] Establishing a main equipment model of the power system, wherein the main equipment model of the power system includes a thermal power unit output power model and an energy storage power station model;

[0114] Output power model of thermal power unit:

[0115] For the unit numbered j, its output at time i during the conventional peak load regulation phase can be expressed as:

[0116] P gen,j,i =P gt,j,i (1+r i ) (1)

[0117] Where: p gt,j,i ——gas turbine power; r i ——Combined cycle steam-combustion power ratio;

[0118] Combined with the actual operation data of thermal power plants, the output of the unit during the start-up and shutdown stages can be expressed by the following piecewise functions (2) and (3) respectively:

[0119]

[0120] Where, t is the time between the start and stop phases of the unit; P gt,j,max ——Rated maximum power of the engine; P gen,j,max ——Maximum power of combined cycle unit, P gen,j,max =P gt,j,max +P wt,j,max ; ——Ramp rate when starting and stopping the gas turbine; ——Ramp rate when the turbine is started and shut down; i a 、i b 、i c 、i d ——Time nodes during the unit startup phase, i e 、i f 、i g ——Shutdown nodes of steam turbines, waste heat boilers and gas turbines;

[0121] Energy storage power station model:

[0122] The charge state of the energy storage power station at time i is as follows:

[0123]

[0124] In the formula, S i ——The charge state of the energy storage power station at time i; Δt——data collection interval; E bn ——Rated capacity of energy storage power station; P bat,cha,i-1 , P bat,dis,i-1 ——The charging and discharging power of the energy storage power station at the previous moment; η bat,cha , η bat,dis ——Charging and discharging efficiency of energy storage power stations.

[0125] Establishing a thermal-solar-storage peak-shaving hierarchical scheduling model includes an upper model and a lower model, determining the objective functions of the upper model and the lower model, and setting the constraints of the upper model and the lower model, including:

[0126] Based on the photovoltaic day-ahead power forecast and the dynamic model of the start-stop peak-shaving unit, an upper-level optimization model is established to optimize the downtime and downtime of the start-stop peak-shaving unit with the goal of minimizing the standard deviation of the equivalent net load fluctuation after the start-stop peak-shaving unit is adjusted; for the unit numbered j, shut down at time x, and with a downtime of y, its output sequence can be temporarily expressed as:

[0127]

[0128] In the formula, i down 、i down,end 、i up 、i up,en ——The time nodes of the unit shutdown and startup process, as shown in formula (6):

[0129]

[0130] The objective function of the upper model is:

[0131]

[0132] Where N T ——Scheduling period; P load,i ——system load at time i; P PV,max,i ——PV power prediction value at time i; P gen,j,i ——The output of peak load regulating unit j started and stopped at time i;

[0133] The constraints of the upper model are:

[0134] Considering that the unit needs to complete the restart action within the remaining time of the scheduling cycle after shutdown, the constraints on the shutdown time and shutdown duration of the unit are:

[0135]

[0136] In the lower model, a multi-energy complementary optimization dispatching model of thermal, solar and storage is established, and the optimization is carried out with the goal of optimizing the economic efficiency of the power system operation within the dispatching period. The objective function is:

[0137] f COST =C gen +C bat +C PV +C net (9)

[0138] In the formula, C gen ——The cost of electricity generation from thermal power plants; C bat ——Operation cost of energy storage power station; C PV ——Penalty for photovoltaic abandonment; C net ——The cost of purchasing electricity from the external grid;

[0139] The power generation cost of a thermal power plant mainly considers the unit fuel cost and start-up and shutdown cost, as shown in the following equations (10) and (11). For the first type of unit, due to its small power regulation range, the fuel cost can be simplified as a linear function of power:

[0140]

[0141]

[0142] Where N gen ——Number of units; C gas,j,i ——Unit fuel cost; δ gen ——Generation cost of the first type of unit; a, b, c——fuel consumption coefficient; δ gas ——Local natural gas selling price; n ss ——Number of unit starts and stops; δ ss ——Single start-up and shutdown cost of the unit;

[0143] The operating cost of the energy storage power station is:

[0144] C bat =(P bat,cha,i +P bat,dis,i )Δtδ bat (12)

[0145] In the formula, δ bat ——Charging and discharging cost coefficient of energy storage;

[0146] The penalty for photovoltaic abandonment is:

[0147]

[0148] Where PPV,loss,i ——the amount of photovoltaic power abandoned at time i; δ PV ——The cost of electricity per kilowatt-hour during the entire life cycle of a photovoltaic power station;

[0149] Cost of purchasing electricity from external network:

[0150]

[0151] Where P net,i ——The power purchased from the external grid at time i; δ net,i ——The electricity price of the State Grid at time i, using the time-of-use electricity price;

[0152] The constraints of the lower model include:

[0153] When the system network loss is ignored, the sum of the thermal, solar, and storage output and purchased electricity of the power system is equal to the real-time load; the power balance constraint of the power system is:

[0154]

[0155] Where P net,i ——The amount of electricity purchased online at time i; P pv,loss,i is the amount of photovoltaic power abandonment at time i;

[0156] For the units in the start-stop peak-shaving stage, their output power is consistent with formula (5); for the units in the rotating peak-shaving stage, their output power is limited by the unit climbing rate and the rated maximum output upper limit; the dynamic constraints of the thermal power plant units are:

[0157]

[0158] Reserve capacity constraints of thermal power plants:

[0159]

[0160] Where U gen,j,i ——The state variable of unit j at time i, 0 represents the unit is shut down, 1 represents the unit is running; Ri——the spare capacity demand of the system at time i;

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

[0162]

[0163] Where P bat,cha,max , P bat,dis,max ——The maximum charging and discharging power of the energy storage station, MW; U cha,i , U dis,i ——Boolean variable for the charge and discharge identification of the energy storage power station, 0 means it is not in the charge or discharge state, and 1 means it is in the charge or discharge state;

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

[0165]

[0166] In the formula, S max , S min ——The upper and lower limits of the state of charge of the energy storage power station; S1, S NT ——Charge status at the beginning and end of the dispatch cycle;

[0167] Restrictions on purchasing electricity from the Internet:

[0168] 0≤P net,i ≤P net,max (20)

[0169] Where P net,max ——The upper limit of electricity purchased from the external grid.

[0170] Determine the solution process of the thermal storage peak-shaving hierarchical scheduling model. The upper model solves the optimal scheduling of the start-stop peak-shaving stage of the unit with the goal of minimizing the equivalent net load fluctuation of the power system; the lower model solves the output scheduling of each unit with the goal of minimizing the operating cost within the power system scheduling cycle; linearize the functions that are inconvenient to be brought into the thermal storage peak-shaving hierarchical scheduling model, including:

[0171] The upper model first obtains the net load characteristics through the photovoltaic day-ahead power prediction model as the input of the upper model; the combined algorithm of FCM and LSTM is used to realize the short-term prediction of photovoltaic power; the long short-term memory neural network, namely LSTM, has a special memory and forgetting mode to flexibly adapt to the temporal characteristics of the learning task; the fuzzy C-means clustering, namely FCM algorithm, can divide the weather type of the sample according to the fluctuation level of the total horizontal irradiance of the historical sample, so as to facilitate the selection of similar days as the training samples of LSTM;

[0172] According to the number of start-stop peak-shaving units put into operation, the particle swarm algorithm is used to minimize the standard deviation of the equivalent net load fluctuation of the power system after the unit adjustment, and the downtime and downtime duration of each start-stop unit are solved cyclically. The equivalent net load adjusted by the previous start-stop peak-shaving unit is used as the input of the next cycle. With the goal of optimizing the operating economy within the power system dispatch cycle, the CPLEX solver is called through Matlab to solve the output dispatch of each unit;

[0173] The following functions that are inconvenient to be brought into the fire storage peak load stratified scheduling model for solution are linearized, including:

[0174] Unit fuel cost:

[0175] First, C gas,j,i Convert to n-segment linear function f(P gt,j,i ), the sub-point is b1,j,i ≤ b 2,j,i ≤…≤b n,j,i ≤ b n+1,j,i , introduce the variable w k,j,i P gt,j,i , and f(P gt,j,i ) is expressed as formula (21), and the constraints of each variable are shown in formula (22):

[0176]

[0177] Energy storage power station charging and discharging power constraints:

[0178] Using the big-M method, the constraints are transformed into:

[0179]

[0180] Where M is a sufficiently large constant.

[0181] The obtained dispatching scheme is implemented in the regional power grid and simulated to monitor the operation status of the power system in real time, and is adjusted and optimized according to the actual situation, including:

[0182] Based on the day-ahead photovoltaic forecast and the start-stop model of the peak-shaving units, the particle swarm algorithm is used to solve the operation and scheduling of the start-stop units of the thermal power plant according to the number of start-stop peak-shaving units put into use; when multiple start-stop units are put into use, the order in which units of different capacities participate in the cycle optimization has an impact on the optimization results. Since the installed capacity of the three units equipped in the power plant is similar, this impact can be ignored, so it is not distinguished. After putting different numbers of start-stop peak-shaving units into use, the equivalent net load results of the power system are solved;

[0183] According to the number of peak load units put into operation, the upper model can generate three optimal dispatching schemes. According to different dispatching schemes, four dispatching scenarios are formed. Dispatching scenario 1 ～ 3 correspond to three dispatching schemes respectively, and dispatching scenario 4 corresponds to not putting into use the peak-shaving units for comparison. These four dispatching scenarios are simulated to obtain the operating cost composition of the power system under various dispatching scenarios.

[0184] Reference Figure 2 :

[0185] Solution process of thermal storage peak load stratified scheduling model: The total number of start-stop peak load units n ss , use the combined algorithm of FCM and LSTM to realize the short-term prediction of photovoltaic power and obtain the photovoltaic predicted power; the number of start-stop peak-shaving units put into regulation is set to i, the initial value i=0, and when looping to solve the downtime and downtime duration (x, y) of the start-stop units, i is increased by 1 each time until i≤n ssIf it is not satisfied, the loop solution is terminated; according to the different number of start-stop peak-shaving units i put into regulation, the equivalent net load curve of the power system is obtained, and the number of start-stop peak-shaving units i put into regulation is increased in sequence. When the number of start-stop peak-shaving units i put into regulation ≤ the number of start-stop peak-shaving units n ss , a loop is performed to solve the objective function using the particle swarm algorithm, and the standard deviation of the equivalent net load fluctuation of the power system after the start-stop peak-shaving units are adjusted is obtained to be the smallest. The number of start-stop peak-shaving units i put into regulation is optimized, and the equivalent net load of the power system after the previous start-stop peak-shaving unit is adjusted is used as the input of the next loop. The downtime and downtime length (x, y) of each start-stop unit and the output of each start-stop peak-shaving unit in the start-stop peak-shaving stage are solved cyclically; when the number of start-stop peak-shaving units i put into regulation ≥ the number of start-stop peak-shaving units n ss When , the loop solution is terminated, and according to the output of each start-stop peak-shaving unit in the start-stop peak-shaving stage, the output of each start-stop peak-shaving unit in the conventional peak-shaving stage and the output of each subsystem are solved by calling CPLEX solver through Matlab with the lowest operating cost within the power system dispatch cycle as the objective function.

[0186] Reference Figure 5 :

[0187] Due to the large proportion of photovoltaic installed capacity, the peak-to-valley difference of the net load of the power system is too high. The more start-stop peak-shaving units are put into use, the flatter the equivalent net load of the power system after adjustment, and the smaller the peak-to-valley difference of the power system; after the three start-stop peak-shaving units are put into use, the overall net load level of the power system is low, basically between 0 and 50MW, thereby compressing the output space of other power sources in the power system; according to the number of start-stop peak-shaving units put into use, the upper model can generate a total of 3 optimal start-stop scheduling schemes. The standard deviation of the original net load of the power system without adjustment is 76.56MW. After the three start-stop peak-shaving units are put into use for adjustment, the standard deviation of the net load of the power system is reduced to 25.30MW at the lowest;

[0188] The results of the start-stop peak load-shaving unit dispatch are shown in Table 1 below:

[0189]

[0190] Reference Figure 6-Figure 9 :

[0191] Analysis of the simulation diagrams of the four dispatching scenarios shows that, on the one hand, the power system mainly completes the regional load supply through thermal power plants during the period without photovoltaic power generation; at the peak of photovoltaic power generation, the power system shuts down some of the first-type units of the thermal power plant to reduce the minimum output of the thermal power plant, cooperates with energy storage equipment to improve the photovoltaic absorption capacity of the power system, and retains the operation of the second-type units to meet the backup rotation capacity requirements of the power system and ensure power supply reliability; on the other hand, during the valley power period, the power system purchases electricity from the external network; during the peak power period, the energy storage power station discharges to reduce the power system's external power purchase;

[0192] In dispatch scenario 2, during the peak period of photovoltaic power generation, the thermal power plant only retains one unit with a low output level, which requires the power system to purchase electricity from the external network to balance the load gap caused by photovoltaic fluctuations; in dispatch scenario 3, the thermal power plant has invested in three start-stop peak-shaving units, resulting in excessive output of the thermal power plant during the period without photovoltaic power generation, which reduces the space for the power system to purchase valley electricity;

[0193] The operating cost composition of the power system under each dispatching scenario is shown in Table 2:

[0194]

[0195] Compared with dispatch scenario 4, after dispatch scenario 1 put into use one start-stop peak-shaving unit, although the start-stop cost of the thermal power plant increased, the power generation of the thermal power plant decreased, and the profit of the thermal power plant was compressed, the abandonment rate of the power system decreased by 11.80%, which reduced the comprehensive operating cost of the power system by 14.14%; dispatch scenario 2 added one start-stop peak-shaving unit on the basis of dispatch scenario 1, and the abandonment rate and operating cost of the power system further decreased by 13.15% and 9.71% respectively; and after too many start-stop peak-shaving units were put into dispatch scenario 3, the abandonment rate and operating cost of the power system were not as good as those of dispatch scenario 2;

[0196] Compared with the single rotating peak-shaving mode, the abandonment rate and operating cost of the power system under the thermal storage peak-shaving stratified dispatching decreased by 34.95% and 22.48% respectively. The results show that the proposed thermal storage peak-shaving stratified dispatching model can effectively improve the photovoltaic absorption level of the regional power grid and reduce the operating cost of the power system.

[0197] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A method for dispatching a photovoltaic thermal energy storage system for a regional power grid, characterized in that: The following steps are involved: Collect the photovoltaic power forecast and load data of the regional power grid, determine the configuration parameters of the power system, and obtain the startup characteristics of the main equipment of the thermal power plant and the units implementing gas-steam combined cycle power generation during the start-stop peak regulation stage; Establishing a main equipment model of the power system, wherein the main equipment model of the power system includes a thermal power unit output power model and an energy storage power station model; Establish a hierarchical dispatching model for peak load regulation of thermal storage, including an upper model and a lower model, determine the objective functions of the upper model and the lower model, and set the constraints of the upper model and the lower model; Determine the solution process of the thermal storage peak-shaving hierarchical scheduling model. The upper model solves the optimal scheduling of the start-stop peak-shaving stage of the unit with the goal of minimizing the equivalent net load fluctuation of the power system; the lower model solves the output scheduling of each unit with the goal of minimizing the operating cost within the power system scheduling cycle; linearize the functions that are inconvenient to be brought into the thermal storage peak-shaving hierarchical scheduling model; The solved dispatching scheme is implemented in the regional power grid and a simulation test is carried out to monitor the operating status of the system in real time, and to make adjustments and optimizations based on actual conditions.

2. A method for dispatching a photovoltaic thermal energy storage system for a regional power grid according to claim 1, characterized in that: Collect the photovoltaic power forecast and load data of the regional power grid, determine the configuration parameters of the power system, and obtain the startup characteristics of the main equipment of the thermal power plant and the units implementing gas-steam combined cycle power generation during the start-stop peak regulation stage, including: Record typical daily photovoltaic and load conditions of regional power grid; The configuration parameters of the power system include 220MW of photovoltaic power generation, 44MW / 88MWh of energy storage power station, and a system spinning reserve capacity of 10% of the load; The main equipment of the thermal power plant is equipped with three first-type gas turbines and one second-type gas turbine; Thermal power plants implement gas-steam combined cycle power generation. For units in the start-stop and peak-shaving stage, the startup characteristics of gas-steam combined cycle power generation during actual operation are divided into three processes: gas turbine startup, grid-connected power generation and reaching maximum load; waste heat boiler startup and pipe warming; steam turbine startup and run-up, grid-connected power generation and reaching maximum load; when the unit is shut down, it is shut down in the order of steam turbine-waste heat boiler-gas turbine.

3. A method for dispatching a photovoltaic thermal energy storage system for a regional power grid according to claim 2, characterized in that: Establishing a main equipment model of the power system, wherein the main equipment model of the power system includes a thermal power unit output power model and an energy storage power station model; Output power model of thermal power unit: For the unit numbered j, its output at time i during the conventional peak load regulation phase can be expressed as: P gen,j,i =P gt,j,i (1+r i ) (1) Where: p gt,j,i ——gas turbine power; r i ——Combined cycle steam-combustion power ratio; Combined with the actual operation data of thermal power plants, the output of the unit during the start-up and shutdown stages can be expressed by the following piecewise functions (2) and (3) respectively: Where, t is the time between the start and stop phases of the unit; P gt,j,max ——Rated maximum power of the engine; P gen,j,max ——Maximum power of combined cycle unit, P gen,j,max =P gt,j,max +P wt,j,max ; ——Ramp rate when starting and stopping the gas turbine; ——Ramp rate when the turbine is started and shut down; i a 、i b 、i c 、i d ——Time nodes during the unit startup phase, i e 、i f 、i g ——Shutdown nodes of steam turbines, waste heat boilers and gas turbines; Energy storage power station model: The charge state of the energy storage power station at time i is as follows: In the formula, S i ——The charge state of the energy storage power station at time i; Δt——data collection interval; E bn ——Rated capacity of energy storage power station; P bat,cha,i-1 , P bat,dis,i-1 ——The charging and discharging power of the energy storage power station at the previous moment; η bat,cha , η bat,dis ——Charging and discharging efficiency of energy storage power stations.

4. A method for dispatching a photovoltaic thermal energy storage system for a regional power grid according to claim 3, characterized in that: Establishing a thermal-solar-storage peak-shaving hierarchical scheduling model includes an upper model and a lower model, determining the objective functions of the upper model and the lower model, and setting the constraints of the upper model and the lower model, including: Based on the photovoltaic day-ahead power forecast and the dynamic model of the start-stop peak-shaving unit, an upper-level optimization model is established to optimize the downtime and downtime of the start-stop peak-shaving unit with the goal of minimizing the standard deviation of the equivalent net load fluctuation after the start-stop peak-shaving unit is adjusted; for the unit numbered j, shut down at time x, and with a downtime of y, its output sequence can be temporarily expressed as: In the formula, i down 、i down,end 、i up 、i up,end ——The time nodes of the unit shutdown and startup process, as shown in formula (6): The objective function of the upper model is: Where N T ——Scheduling period; P load,i ——system load at time i; P PV,max,i ——PV power prediction value at time i; P gen,j,i ——The output of peak load regulating unit j started and stopped at time i; The constraints of the upper model are: Considering that the unit needs to complete the restart action within the remaining time of the scheduling cycle after shutdown, the constraints on the shutdown time and shutdown duration of the unit are: In the lower model, a multi-energy complementary optimization dispatching model of thermal, solar and storage is established, and the optimization is carried out with the goal of optimizing the economic efficiency of the power system operation within the dispatching period. Its objective function is: f COST =C gen +C bat +C PV +C net (9) In the formula, C gen ——The cost of electricity generation from thermal power plants; C bat ——Operation cost of energy storage power station; C PV ——Penalty for photovoltaic abandonment; C net ——The cost of purchasing electricity from the external grid; The power generation cost of a thermal power plant mainly considers the unit fuel cost and start-up and shutdown cost, as shown in the following equations (10) and (11). For the first type of unit, due to its small power regulation range, the fuel cost can be simplified as a linear function of power: Where N gen ——Number of units; C gas,j,i ——Unit fuel cost; δ gen ——Generation cost of 6B type unit; a, b, c——fuel consumption coefficient; δ gas ——Local natural gas selling price; n ss ——Number of unit starts and stops; δ ss ——Single start-up and shutdown cost of the unit; The operating cost of the energy storage power station is: C bat =(P bat,cha,i +P bat,dis,i )Δtδ bat (12) In the formula, δ bat ——Charging and discharging cost coefficient of energy storage; The penalty for photovoltaic abandonment is: Where P PV,loss,i ——the amount of photovoltaic power abandoned at time i; δ PV ——The cost of electricity per kilowatt-hour during the entire life cycle of a photovoltaic power station; Cost of purchasing electricity from external network: Where P net,i ——The power purchased from the external grid at time i; δ net,i ——The electricity price of the State Grid at time i, using the time-of-use electricity price; The constraints of the lower model include: When the system network loss is ignored, the sum of the thermal, solar, and storage output and purchased electricity of the power system is equal to the real-time load; the power balance constraint of the power system is: Where P net,i ——The amount of electricity purchased online at time i; P pv,loss,i is the amount of photovoltaic power abandonment at time i; For the units in the start-stop peak-shaving stage, their output power is consistent with formula (5); for the units in the rotating peak-shaving stage, their output power is limited by the unit climbing rate and the rated maximum output upper limit; the dynamic constraints of the thermal power plant units are: Reserve capacity constraints of thermal power plants: Where U gen,j,i ——The state variable of unit j at time i, 0 represents the unit is shut down, 1 represents the unit is running; Ri——the spare capacity demand of the system at time i; Energy storage power station charging and discharging power constraints: Where P bat,cha,max , P bat,dis,max ——The maximum charging and discharging power of the energy storage station, MW; U cha,i , U dis,i ——Boolean variable for charging and discharging identification of the energy storage power station, 0 means it is not in the charging or discharging state, and 1 means it is in the charging or discharging state; State of charge constraints of energy storage power stations: In the formula, S max , S min ——The upper and lower limits of the state of charge of the energy storage power station; S1, S NT ——Charge status at the beginning and end of the dispatch cycle; Restrictions on purchasing electricity from the Internet: 0≤P net,i ≤P net,max (20) Where P net,max ——The upper limit of electricity purchased from the external grid.

5. A method for dispatching a photovoltaic thermal energy storage system for a regional power grid according to claim 4, characterized in that: Determine the solution process of the thermal storage peak-shaving hierarchical scheduling model. The upper model solves the optimal scheduling of the start-stop peak-shaving stage of the unit with the goal of minimizing the equivalent net load fluctuation of the power system; the lower model solves the output scheduling of each unit with the goal of minimizing the operating cost within the power system scheduling cycle; linearize the functions that are inconvenient to be brought into the thermal storage peak-shaving hierarchical scheduling model, including: The upper model first obtains the net load characteristics through the photovoltaic day-ahead power prediction model as the input of the upper model; the combined algorithm of FCM and LSTM is used to realize the short-term prediction of photovoltaic power; According to the number of start-stop peak-shaving units put into operation, the particle swarm algorithm is used to minimize the standard deviation of the equivalent net load fluctuation of the power system after the unit adjustment, and the downtime and downtime duration of each start-stop unit are solved cyclically. The equivalent net load adjusted by the previous start-stop peak-shaving unit is used as the input of the next cycle. With the goal of optimizing the operating economy within the power system dispatch cycle, the CPLEX solver is called through Matlab to solve the output dispatch of each unit; The following functions that are inconvenient to be brought into the fire storage peak load stratified scheduling model for solution are linearized, including: Unit fuel cost: First, C gas,j,i Convert to n-segment linear function f(P gt,j,i ), the sub-point is b 1,j,i ≤ b 2,j,i ≤…≤b n,j,i ≤ b n+1,j,i , introduce the variable w k,j,i P gt,j,i , and f(P gt,j,i ) is expressed as formula (21), and the constraints of each variable are shown in formula (22): Energy storage power station charging and discharging power constraints: Using the big-M method, the constraints are transformed into: Where M is a sufficiently large constant.

6. A method for dispatching a photovoltaic thermal energy storage system for a regional power grid according to claim 5, characterized in that: The obtained dispatching scheme is implemented in the regional power grid and simulated to monitor the operation status of the power system in real time, and is adjusted and optimized according to the actual situation, including: Based on the day-ahead photovoltaic forecast and the start-stop model of the peak-shaving units, the particle swarm algorithm is used to solve the operation and dispatch of the start-stop units of the thermal power plant according to the number of start-stop peak-shaving units put into use; after putting in different numbers of start-stop peak-shaving units, the equivalent net load results of the power system are solved; According to the number of peak-shaving units put into operation, the upper-level model can generate three optimal dispatching schemes. According to different dispatching schemes, four dispatching scenarios are formed. These four dispatching scenarios are simulated to obtain the operating cost composition of the power system under various dispatching scenarios.

Images