Photovoltaic photo-thermal multi-energy complementary system automatic power generation control method

Abstract

The application discloses a kind of photovoltaic photothermal multi-energy complementary system automatic power generation control method, establishes the dynamic model comprising photovoltaic power station, photothermal power station, heat storage system, frequency response, tie-line exchange power and model predictive controller;Frequency deviation, tie-line power deviation, each power station output, heat storage and load disturbance are collected, and discrete state space prediction model is constructed.To regional control deviation, frequency and tie-line deviation, heat storage state deviation and control amount change as optimization goal, under photovoltaic reserve margin, photothermal output and climbing, heat storage capacity, active balance and frequency safety and other constraints, rolling solution photovoltaic, photothermal optimal control amount.Only first control amount is issued every cycle, and according to actual and predicted output deviation, correct next cycle initial value.The application does not need external prediction data, can coordinate photovoltaic fast response and photothermal sustained support, reduce frequency deviation and overshoot, shorten recovery time, improve system stability and automatic power generation control precision.

Description

Technical Field

[0001] This invention relates to the field of new energy power generation control and automatic power generation control technology, and in particular to an automatic power generation control method for a photovoltaic-thermal multi-energy complementary system. Background Technology

[0002] In a photovoltaic-thermal complementary system, although both photovoltaic (PV) and solar thermal (CSP) power plants can participate in automatic power generation control, their constraint mechanisms are different. PV power plants connect to the grid via power electronic devices, offering the advantage of fast response speed, but their adjustability is not a fixed value. Instead, it is limited by current active power output, grid connection control limits, and the instantaneously adjustable dynamic reserve margin. CSP power plants, on the other hand, rely on thermal storage systems and thermoelectric conversion processes to participate in frequency regulation. Their continuous support capacity is closely related to the state of thermal storage, the unit ramp-up rate, and the requirements for maintaining end-point thermal storage.

[0003] Most existing automatic control methods for photovoltaic (PV) and solar thermal power generation first generate the total system control quantity based on frequency deviation, and then allocate it proportionally or with static weights. The shortcoming of this approach is not whether the control logic is closed-loop, but rather that the control quantity is allocated secondaryly after the total quantity is generated. This results in the control command formation stage not explicitly considering the dynamic reserve margin of PV and the constraints of thermal storage status. In actual operation, this approach easily leads to three types of problems: First, when the PV power plant's current output is already close to its adjustable upper or lower limit, it is still allocated a control quantity exceeding its dynamic adjustment capacity, resulting in insufficient executability of control commands; second, during frequency recovery, the solar thermal power plant may rapidly consume thermal storage energy due to excessive compensation tasks, weakening its subsequent continuous support capacity; third, when frequency deviation, tie-line exchange power deviation, and thermal storage status constraints coexist, the allocation method based on the total control quantity makes it difficult to coordinate multiple constraint objectives within a unified framework, easily leading to conflicts between frequency recovery, thermal storage maintenance, and control smoothness.

[0004] Therefore, it is necessary to propose an automatic power generation control method for photovoltaic and solar thermal multi-energy complementary systems. This method directly considers the dynamic reserve margin of photovoltaics, the ramp-up capability of solar thermal units, and the capacity status of the thermal storage system during the optimization solution stage. This ensures that the control variables meet multiple operating boundaries from the generation stage, thereby improving the executability of control commands and the continuous support capability of thermal storage. Summary of the Invention

[0005] The purpose of this invention is to provide an automatic power generation control method for a photovoltaic-thermal multi-energy complementary system, in order to solve the problem that existing automatic power generation control methods for photovoltaic and solar thermal systems cannot simultaneously take into account the dynamic reserve margin of photovoltaic power plants, the ramp-up constraints of solar thermal power plants, the capacity constraints of thermal storage systems, and the power exchange constraints of tie lines, resulting in large system frequency fluctuations, high overshoot, and long stabilization times.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: An automatic power generation control method for a photovoltaic-thermal multi-energy complementary system includes the following steps: S1. Establish an automatic power generation control dynamic model for a photovoltaic-thermal multi-energy complementary system, including a photovoltaic power plant, a solar thermal power plant, a thermal storage system, a system frequency response component, a tie-line power exchange component, and a model predictive controller. S2, collect system frequency deviation, tie line switching power deviation, current output power of photovoltaic power station, current output power of solar thermal power station, current heat storage of thermal storage system and load disturbance change, and construct regional control deviation; S3. Based on the dynamic response relationship between photovoltaic power plants and solar thermal power plants, establish a discrete state-space prediction model;

[0007] in, For system state variables, To control variables, For the disturbance variable, For output variables, , , , These are the discrete state matrix, control input matrix, disturbance matrix, and output matrix, respectively. S4. Set the prediction time domain N p and control time domain N c Construct the model predictive control rolling optimization objective function:

[0008] in ; S5. Under the constraints of dynamic reserve margin of photovoltaic power plant, output of solar thermal power plant, ramp-up rate of solar thermal unit, capacity of thermal storage system, active power balance of system, frequency security and power deviation of tie line, solve the optimal control sequence in the future control time domain. S6. Take the first control quantity in the optimal control sequence and apply it to the photovoltaic power station and the solar thermal power station; S7. In the next control cycle, the initial state of the prediction model is corrected by feedback based on the deviation between the actual output of the system and the predicted output of the previous control cycle, and S2 to S6 are repeated to realize the closed-loop automatic power generation control of the photovoltaic and solar thermal multi-energy complementary system.

[0009] As a preferred embodiment of the present invention, the regional control deviation is constructed using a tie-line power frequency deviation control method:

[0010] In the formula, This is the frequency deviation coefficient. This is the system frequency deviation. This refers to the power deviation of the tie line.

[0011] As a preferred embodiment of the present invention, in step S3, the system state variable is:

[0012] In the formula, For the increase in output power of photovoltaic power plants, For the increase in output power of solar thermal power plants, For storing heat in the thermal storage system, For the valve position increment of the solar thermal unit speed controller; The control variable is:

[0013] in, The control quantity is issued to the photovoltaic power station. This is the control quantity issued to the solar thermal power plant.

[0014] As a preferred embodiment of the present invention, the dynamic response model of the photovoltaic power station is as follows: in, The delay constant of photovoltaic AGC. The inverter time constant, The photovoltaic power response time constant is For photovoltaic power response gain; The dynamic reserve margin constraint of the photovoltaic power station is:

[0015] in, For the first k The dynamic reserve margin that can be increased by the photovoltaic power station at any time. For the first k The dynamic reserve margin that the photovoltaic power station can use for downward adjustment at any time; the dynamic reserve margin is determined based on the current active power output of the photovoltaic power station, the dispatchable output range, and the grid connection control limit.

[0016] As a preferred embodiment of the present invention, the adjustable dynamic reserve margin and adjustable dynamic adjustment margin of the photovoltaic power station are respectively:

[0017] In the formula, The active power output of the photovoltaic power station at time k is... This represents the current maximum active power capacity allowed for automatic generation control in photovoltaic power plants. This represents the current minimum active power allowed for photovoltaic power plants to participate in automatic power generation control.

[0018] As a preferred embodiment of the present invention, the active power response process of the solar thermal power plant includes AGC delay, governor, reheat turbine, and generator power response stages; the control command after the AGC delay is:

[0019] In the formula, s For the Laplace operator, The delay constant for AGC commands in a solar thermal power plant. The controller sends control signals to the solar thermal power plant. The control command is the result of the AGC delay period. The dynamic relationship of the speed governor can be expressed as:

[0020] In the formula, For the incremental position of the governor valve of the solar thermal unit. For speed controller gain, The time constant of the speed controller, R This is the unit droop coefficient. This refers to the system frequency deviation. The mechanical power output of the reheat turbine stage is:

[0021] In the formula, For the increase in mechanical power of the steam turbine, For turbine power gain, The time constant of the steam turbine. The reheat coefficient, The reheat time constant; The output power increment of the solar thermal power plant meets the following requirements:

[0022] In the formula, For the increase in output power of solar thermal power plants, Let be the power response time constant of the solar thermal generator. Based on the above dynamic response model of the solar thermal power plant; The output constraint of the solar thermal power plant is:

[0023] The ramp rate constraint for the solar thermal unit is:

[0024] In the formula, and These represent the minimum and maximum active power output of a solar thermal power plant, respectively. and These are the maximum downhill and uphill speeds of the solar thermal power unit, respectively.

[0025] As a preferred embodiment of the present invention, the energy state of the thermal storage system satisfies:

[0026] In the formula, For the first k The equivalent heat charging power that constantly enters the thermal storage system. To supply thermal power to the solar thermal power generation modules of the thermal storage system, For thermoelectric conversion efficiency, This is the heat storage loss coefficient.

[0027] As a preferred embodiment of the present invention, the capacity constraint of the thermal storage system includes upper and lower limits of thermal storage capacity and end-point thermal storage retention constraint:

[0028] In the formula, and These represent the minimum and maximum heat storage capacity of the thermal storage system, respectively. To predict the lower limit of heat storage at the end of the time domain.

[0029] As a preferred embodiment of the present invention, the system frequency dynamic equation and the tie-line dynamic equation are as follows:

[0030] in, H The system's equivalent inertial time constant. D This is the load damping coefficient. This is the synchronization coefficient for the tie line; The active power balance constraint of the system adopts an incremental form:

[0031] In the formula, This represents the change in load disturbance. For power balance slack variables, For the increase in output power of photovoltaic power plants, For the increase in output power of solar thermal power plants, This is to compensate for the power deviation of the tie line.

[0032] As a preferred embodiment of the present invention, the feedback correction includes:

[0033] In the formula, For prediction error, For the first k Actual measured output at time +1 For the first k Time for the first k Predicted output at time +1 For feedback correction gain matrix, This is the initial value predicted for the next control cycle after feedback correction.

[0034] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention does not adopt the method of first generating the total control quantity of the system and then performing secondary allocation. Instead, it directly solves the control quantities of photovoltaic power plants and solar thermal power plants under the model predictive control framework, so that the control commands meet the dynamic reserve margin, thermal storage capacity and ramp-up constraints at the generation stage.

[0035] 2. This invention incorporates both the dynamic reserve margin of photovoltaic power and the end-point maintenance constraint of the thermal storage system into the automatic power generation control process, which can prevent the photovoltaic control quantity from exceeding the instantaneous adjustable capacity and the solar thermal power plant from excessively consuming thermal storage energy during short-term frequency regulation.

[0036] 3. This invention achieves a unified optimization of photovoltaic rapid response capability and solar thermal continuous support capability, which is conducive to improving the executability of control commands and the subsequent continuous frequency regulation capability of solar thermal power plants. Attached Figure Description

[0037] To more clearly illustrate the implementation of the present invention or the existing technical solutions, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0038] Figure 1 This is a diagram illustrating the overall architecture of the automatic power generation control system for a photovoltaic-thermal multi-energy complementary system provided in an embodiment of the present invention. Figure 2 A diagram illustrating the dynamic reserve margin constraint model for a photovoltaic power plant provided in this embodiment of the invention; Figure 3 A constraint model diagram of a solar thermal power plant and thermal storage system provided in an embodiment of the present invention; Figure 4 This is a flowchart of the MPC rolling optimization control provided in an embodiment of the present invention; Figure 5 A feedback correction flowchart provided for embodiments of the present invention; Figure 6 A comparison chart of the photovoltaic-thermal synergistic frequency modulation control effects provided in this embodiment of the invention. Detailed Implementation

[0039] The automatic power generation control method for a photovoltaic-thermal multi-energy complementary system according to the present invention will be described in complete and detailed manner below with reference to the accompanying drawings. This section fully elaborates on the implementation process of the present invention through the overall system architecture, functional module composition, dynamic model establishment, constraint setting, rolling optimization control process, and simulation effect verification, so that those skilled in the art can clearly and accurately understand and reproduce the technical solution of the present invention. This embodiment is only used to illustrate the present invention and is not intended to limit the scope of protection of the present invention.

[0040] This embodiment discloses an automatic power generation control method for a photovoltaic-thermal multi-energy complementary system based on MPC. This method is implemented based on the automatic power generation control system of the photovoltaic-thermal multi-energy complementary system. For example... Figure 1 As shown, the system uses photovoltaic power plants, solar thermal power plants, and thermal storage systems as core power generation and energy storage units, system frequency response links and tie-line power exchange links as grid operation interfaces, and model predictive controllers as the unified control decision core, forming a complete closed-loop automatic power generation and control system.

[0041] From an overall structural perspective, this system mainly consists of a photovoltaic power plant module, a solar thermal power plant module, a thermal storage system module, a system frequency and tie-line power feedback module, and a model predictive controller module. These modules form a closed-loop control link through status acquisition, control command transmission, power response, and feedback correction, collectively achieving high-precision automatic power generation control of the photovoltaic-solar thermal multi-energy complementary system.

[0042] To uniformly describe the state coupling relationships between modules, the system state variables are defined as follows: (1) The control variables are: (2) The disturbance variable is: (3) The output variable is: (4) In the formula, For system frequency deviation, For the offset of the switching power of the tie line, For the increase in output power of photovoltaic power plants, For the increase in output power of solar thermal power plants, For storing heat in the thermal storage system, For the incremental position of the governor valve of the solar thermal unit. For photovoltaic power plant control quantities, For control parameters of solar thermal power plants This represents the change in load disturbance.

[0043] Based on the above variable definitions, the closed-loop automatic power generation control model of the photovoltaic-thermal multi-energy complementary system can be expressed as: (5) (6) In the formula, , , , These are the discrete state matrix, control input matrix, disturbance matrix, and output matrix, respectively. The model predictive controller predicts the future changes in the system state in the time domain based on equations (5) and (6), and solves for the optimal control quantities of the photovoltaic power station and the solar thermal power station under multiple operating constraints.

[0044] The first part is the photovoltaic power station module. This module operates in grid connection via an inverter and mainly includes an AGC delay stage, an inverter response stage, and a photovoltaic power response stage. Photovoltaic power stations are characterized by fast adjustment speed and rapid command execution, making them suitable for quickly providing active power support and suppressing rapid system frequency shifts during initial stages of load disturbances, tie-line power deviations, or sudden changes in photovoltaic output. Simultaneously, to ensure the executable execution of control commands, this invention does not use a fixed proportional allocation of control quantities. Instead, it determines a dynamic reserve margin based on the real-time operating status, ensuring that control commands always remain within the currently adjustable range, thus preventing command overreach and execution failures from the outset.

[0045] After receiving the power control command from the model predictive controller, the photovoltaic power station sequentially completes the grid-connected active power regulation through the AGC command delay stage, the inverter response stage, and the power response stage. Its dynamic response model is expressed as follows: (7) In the formula, For the increase in output power of photovoltaic power plants, The model predicts the control quantities that the controller will send to the photovoltaic power station. For photovoltaic power response gain, The delay constant of photovoltaic AGC. The inverter time constant, is the photovoltaic power response time constant.

[0046] To fundamentally ensure the executability of control commands, this invention collects the operating status of the photovoltaic power station in real time during each control cycle and dynamically calculates the upward and downward adjustment margins under the current operating conditions. k The control parameters of the photovoltaic power plant at any given time must meet the following dynamic reserve margin constraints: (8) in, For the first kThe dynamic reserve margin that can be increased by the photovoltaic power station at any time. For the first k The dynamic adjustment margin that can be used to adjust photovoltaic power stations at any given time. (9) (10) In the formula, The active power output of the photovoltaic power station at time k is... This represents the current maximum active power capacity allowed for automatic generation control in photovoltaic power plants. This represents the current minimum active power allowed for photovoltaic power plants to participate in automatic power generation control.

[0047] The second part is the concentrated solar power (CSP) plant module. A CSP plant provides heat through a thermal storage system, performs work through a reheat turbine, and outputs electrical energy through a generator. Its dynamic response process sequentially includes the AGC command delay stage, the governor stage, the reheat turbine stage, and the generator power response stage. Compared to photovoltaic (PV) power plants, the CSP plant's response process is relatively smooth, but it possesses continuous and stable power support capabilities, making it suitable for undertaking the main continuous regulation tasks during system frequency recovery phases, ensuring long-term system power balance and grid operation stability.

[0048] After receiving control commands from the model predictive controller, the concentrated solar power (CSP) plant first goes through an AGC command delay stage. Let the control quantity sent from the model predictive controller to the CSP plant be... The control command after the AGC delay is Then there is (11) In the formula, s For the Laplace operator, The delay constant for AGC commands in a solar thermal power plant. The controller sends control signals to the solar thermal power plant. This is the control command after the AGC delay stage.

[0049] The speed controller is used to adjust the valve opening according to AGC control commands and frequency feedback. Its dynamic relationship can be expressed as follows: (12) In the formula, For the incremental position of the governor valve of the solar thermal unit. For speed controller gain, The time constant of the speed controller, R This is the unit droop coefficient. This represents the system frequency deviation.

[0050] The reheat turbine stage is used to characterize the dynamic process of converting thermal power into mechanical power in a solar thermal power unit. Its output mechanical power increment can be expressed as: (13) In the formula, For the increase in mechanical power of the steam turbine, For turbine power gain, The time constant of the steam turbine. The reheat coefficient, is the reheat time constant.

[0051] The generator power response element characterizes the conversion process from mechanical power to electrical power output. The power increment of a solar thermal power plant satisfies the following: (14) In the formula, For the increase in output power of solar thermal power plants, Let be the power response time constant of the solar thermal generator. Using the above dynamic response model of the solar thermal power plant, the MPC control quantity can be... Speed ​​controller valve position increment and the increase in output power of solar thermal power plants Establishing connections provides a dynamic foundation for the subsequent construction of the state-space model.

[0052] The third part is the thermal energy storage system (TES). The TES is a core supporting unit for the continuous regulation of a concentrated solar power (CSP) plant, responsible for storing and releasing thermal energy, and is crucial to ensuring the CSP plant's frequency regulation endurance. To avoid excessive consumption of stored thermal energy and weakening the system's subsequent regulation capabilities during short-term frequency regulation, this invention sets dual constraints on the TES: upper and lower limits on stored thermal capacity and a predicted time-domain end-of-pipe stored thermal capacity retention constraint. This ensures that the CSP plant retains necessary subsequent regulation margins while participating in short-term frequency regulation, balancing immediate regulation performance with continuous operation capability.

[0053] The energy state of the thermal storage system changes in real time with the charging power, the releasing power, and the inherent heat loss. Its energy state update equation is: (15) In the formula, For the first k The heat storage system stores heat at all times. For the first k The equivalent heat charging power that constantly enters the thermal storage system. To supply thermal power to the solar thermal power generation modules of the thermal storage system, For thermoelectric conversion efficiency, This is the heat storage loss coefficient.

[0054] The electrical output power of the solar thermal power plant and the thermal power supplied to the power generation module satisfy the following: (16) In the formula, For the first k The power output of the solar thermal power plant at any time This refers to the thermoelectric conversion efficiency.

[0055] To avoid overcharging or over-discharging of the thermal storage system, the capacity constraint of the thermal storage system is as follows: (17) Meanwhile, to avoid excessive consumption of thermal energy storage during short-term frequency regulation and its impact on subsequent support capabilities, this invention sets a constraint on the retention of thermal energy storage at the end of the predicted time domain: (18) In the formula, and These represent the minimum and maximum heat storage capacity of the thermal storage system, respectively. To predict the lower limit of heat storage at the end of the time domain.

[0056] In terms of unit operation safety constraints, to avoid the problem of long-term over-rated or low-output inefficient operation of concentrated solar power (CSP) units, the active power output of the CSP plant must meet the following requirements: (19) Meanwhile, to prevent safety hazards such as equipment vibration, thermal stress shock, and pipeline pressure fluctuations caused by large and sudden power changes per unit time, the ramp rate of the solar thermal power unit must meet the following requirements: (20) In the formula, and These represent the minimum and maximum active power output of a solar thermal power plant, respectively. and These are the maximum downhill ramp rate and maximum uphill ramp rate of the solar thermal power unit, respectively. The fourth part is the system frequency and tie-line power component. This component is used to collect two core power grid operating parameters in real time: system frequency deviation and tie-line switching power deviation. The collected parameter signals are then sent to the model predictive controller to construct regional control deviations, thereby achieving the coordinated control objective of power grid frequency security and stability and controllable tie-line switching power.

[0057] The system frequency deviation is defined as: (twenty one) In the formula, For the first k Real-time system measured frequency This is the system's rated frequency.

[0058] The area control deviation is constructed using the tie-line power-frequency deviation control method, and its expression is: (twenty two) In the formula, For regional control deviation, This is the frequency deviation coefficient. This is the system frequency deviation. This refers to the power deviation of the tie line.

[0059] The system frequency dynamic equation and tie-line dynamic equation are as follows: (twenty three) (twenty four) in, H The system's equivalent inertial time constant. D This is the load damping coefficient. This is the synchronization coefficient for the tie line.

[0060] The fifth part is the Model Predictive Controller (MPC). As the decision-making core of the entire system, the MPC coordinates the entire closed-loop control process, including state acquisition, predictive model construction, rolling optimization solution, control command issuance, and feedback correction. Unlike the traditional method of first generating the total system control quantity and then performing secondary proportional allocation, the MPC in this invention directly solves for the optimal control quantities of the photovoltaic power station and the solar thermal power station, enabling the control commands to be adapted to the operating constraints of each power generation unit during the generation stage, thereby improving the rationality, executability, and closed-loop regulation stability of the system.

[0061] The model predictive controller solves for the following control sequence in the prediction time domain: (25) Among them, the single-period control variable is: (26) The model predictive controller takes area control deviation, frequency deviation, tie-line exchange power deviation, thermal storage state deviation, and the magnitude of control variable changes as comprehensive optimization objectives. Its objective function is expressed as: (27) The change in the control quantity is: (28) In the formula, , , , These are the weighting coefficients for regional control deviation, frequency deviation, tie-line exchange power deviation, and thermal storage status deviation, respectively. R For the control quantity weight matrix;S For the control quantity change rate weight matrix; p ε The penalty coefficient for slack variables; e These are slack variables.

[0062] Through the coordinated operation of the above modules, this system can fully leverage the advantages of photovoltaic's rapid response and instantaneous fluctuation mitigation, as well as the advantages of solar thermal's continuous support and long-term stable regulation. It solves the problems existing in traditional photovoltaic-solar thermal coordinated control, such as unreasonable allocation of control commands, poor adaptability of unit constraints, large frequency fluctuation overshoot, excessive consumption of thermal storage resources, and insufficient system stability. It achieves high-precision and highly robust automatic power generation closed-loop control of photovoltaic-solar thermal multi-energy complementary power generation system.

[0063] In a photovoltaic-thermal multi-energy complementary grid-connected power generation system, the photovoltaic power station relies on power electronic equipment to achieve power regulation, exhibiting excellent response speed. However, due to the influence of real-time grid-connected output, dispatch operation limits, and instantaneous operating conditions, its adjustable power range is not a fixed constant, and the dynamic adjustment margin has time-varying characteristics. If the traditional fixed-proportion power allocation method is used, it is very easy for the allocation control quantity to exceed the real-time adjustable range of the photovoltaic system, leading to problems such as control command exceeding limits, execution failure, and system regulation lag, which seriously affects the grid frequency regulation effect. To address this, this embodiment specifically builds a full-link dynamic response model for the photovoltaic power station and introduces an adaptive dynamic reserve margin constraint mechanism to accurately characterize the real-time regulation capability of the photovoltaic power station. The corresponding constraint model structure is as follows: Figure 2 As shown.

[0064] After receiving power control commands from the model predictive controller, a photovoltaic power plant cannot instantly complete a power response. It must sequentially pass through three dynamic links: the AGC command delay link, the inverter response link, and the power inertia response link, to gradually adjust the grid-connected active power. The entire process can be characterized by a series of multi-stage inertial links, fully reconstructing the timing and gain characteristics of photovoltaic command transmission, equipment response, and power output. Its dynamic response model is expressed as follows: (29) In the formula, For the increase in output power of photovoltaic power plants, The model predicts the control quantities that the controller will send to the photovoltaic power station. For photovoltaic power response gain, The delay constant of photovoltaic AGC. The inverter time constant, is the photovoltaic power response time constant.

[0065] To fundamentally ensure the executability of control commands, this invention abandons the fixed adjustment range setting method and collects photovoltaic power station operating parameters in real time during each control cycle, dynamically calculating the upward and downward adjustment margins under the current operating conditions. Specifically, in the first... k In each control cycle, by collecting the current actual active power output of the photovoltaic power station and the upper and lower limits of active power adjustment authorized by the dispatching AGC, the instantaneous adjustable range of the photovoltaic power station is quantified, providing accurate boundary constraints for subsequent optimization solutions. k The control parameters of the photovoltaic power plant at any given time must meet the following dynamic reserve margin constraints: (30) Among them, the adjustable dynamic reserve margin represents the maximum active power that the photovoltaic power station can generate under the current operating conditions, and the adjustable dynamic adjustment margin represents the maximum active power that the photovoltaic power station can reduce under the current operating conditions. Their calculation formulas are as follows: (31) (32) In the formula, The active power output of the photovoltaic power station at time k is... This represents the current maximum active power capacity allowed for automatic generation control in photovoltaic power plants. This represents the current minimum active power allowed for photovoltaic power plants to participate in automatic power generation control.

[0066] In each round of rolling optimization, the model predictive controller forces the control quantities of the photovoltaic power station to be constrained within the margin range calculated in real time, avoiding the problem of mismatch between control commands and the actual adjustment capabilities of the equipment. This effectively improves the accuracy and executability of control commands and ensures that the photovoltaic power station can participate in system frequency regulation efficiently and stably.

[0067] Compared to photovoltaic power plants, concentrated solar power (CSP) power plants exhibit power regulation characteristics highly dependent on the energy storage status of the thermal storage system. Furthermore, the units themselves are constrained by equipment safety operation limitations, with fixed upper and lower output limits and ramp-up rates. Their regulation characteristics combine thermal inertia and mechanical constraints. Ignoring the operational boundaries of the CSP units and thermal storage systems can easily lead to problems such as units operating beyond their capacity, sudden power fluctuations impacting equipment, overdraft of thermal storage energy, and loss of subsequent frequency regulation capabilities. To ensure the safe, stable, and continuous participation of the CSP power plant in the system's automatic power generation control, this embodiment constructs a complete dynamic response model for the CSP power plant and a refined energy management model for the thermal storage system, comprehensively constraining the system's operating status. The corresponding constraint model structure is as follows: Figure 3 As shown.

[0068] The power response of a solar thermal power plant is a slow dynamic process involving multiple series stages. After receiving control commands from the MPC controller, the command first passes through the AGC command delay stage to complete command transmission and delay correction. Subsequently, the command is input to the speed governor stage, which, in conjunction with the system frequency deviation feedback signal, adjusts the turbine valve opening in real time to control the steam flow into the reheat turbine. The reheat turbine completes the conversion of thermal energy into mechanical power, and finally, the generator converts the mechanical power into grid-connected electrical power, thus fully realizing the dynamic response process from control command to active power output.

[0069] After receiving control commands from the model predictive controller, the concentrated solar power (CSP) plant first goes through an AGC command delay stage. Let the control quantity sent from the model predictive controller to the CSP plant be... The control command after the AGC delay is Then we have: (33) In the formula, s For the Laplace operator, The delay constant for AGC commands in a solar thermal power plant. The controller sends control signals to the solar thermal power plant. This is the control command after the AGC delay stage.

[0070] The speed controller is used to adjust the valve opening according to AGC control commands and frequency feedback. Its dynamic relationship can be expressed as follows: (34) In the formula, For the incremental position of the governor valve of the solar thermal unit. For speed controller gain, The time constant of the speed controller, R This is the unit droop coefficient. This represents the system frequency deviation.

[0071] The reheat turbine stage is used to characterize the dynamic process of converting thermal power into mechanical power in a solar thermal power unit. Its output mechanical power increment can be expressed as: (35) In the formula, For the increase in mechanical power of the steam turbine, For turbine power gain, The time constant of the steam turbine. The reheat coefficient, is the reheat time constant.

[0072] The generator power response element characterizes the conversion process from mechanical power to electrical power output. The power increment of a solar thermal power plant satisfies the following: (36) In the formula, For the increase in output power of solar thermal power plants, is the power response time constant of the solar thermal generator.

[0073] In terms of unit operation safety constraints, to avoid the problem of long-term operation of solar thermal power units exceeding rated output or operating at low output and inefficiency, upper and lower limits are set for the active power output of the units to restrict the actual operating output range of the units: (37) Meanwhile, to prevent safety hazards such as equipment vibration, thermal stress shock, and pipeline pressure fluctuations caused by large and sudden power changes within a unit time, unit ramp rate constraints are set to strictly limit the maximum rate of change of unit power increase and decrease, ensuring the safe and stable operation of the solar thermal power unit equipment: (38) In the formula, and These represent the minimum and maximum active power output of a solar thermal power plant, respectively. and These are the maximum downhill ramp rate and maximum uphill ramp rate of the solar thermal power unit, respectively. As the core guarantee for continuous frequency regulation in a solar thermal power plant, the thermal storage system's stored heat capacity changes dynamically in real time during the system's charging and releasing processes. The remaining thermal storage capacity in each control cycle is determined by the remaining stored heat capacity from the previous cycle, the current charging power, the power generation and releasing power, and the system's inherent heat losses. The energy state update equation for the thermal storage system is: (39) In the formula, For the first k The heat storage system stores heat at all times. For the first k The equivalent heat charging power that constantly enters the thermal storage system. To supply thermal power to the solar thermal power generation modules of the thermal storage system, For thermoelectric conversion efficiency, This is the heat storage loss coefficient.

[0074] The electrical output power of the solar thermal power plant and the thermal power supplied to the power generation module satisfy the following: (40) In the formula, For the first k The power output of the solar thermal power plant at any time This refers to the thermoelectric conversion efficiency.

[0075] To address the shortcomings of traditional control strategies, such as excessive consumption of thermal energy storage during short-term frequency regulation leading to insufficient adjustment margin for subsequent load fluctuations, this invention establishes a dual thermal energy storage constraint mechanism. This mechanism not only limits the maximum and minimum heat storage capacity of the thermal energy storage system to prevent overcharging and over-discharging, but also adds a constraint on thermal energy storage retention at the end of the predictive time domain. This ensures that the thermal energy storage system retains sufficient energy storage margin after the end of the current predictive control cycle, guaranteeing the solar thermal power plant's continuous and long-term grid support capability, while balancing short-term frequency regulation accuracy with long-term operational stability. The thermal energy storage system capacity constraint is as follows: (41) The predicted time-domain end-of-time thermal storage retention constraint is: (42) In the formula, and These represent the minimum and maximum heat storage capacity of the thermal storage system, respectively. To predict the lower limit of heat storage at the end of the time domain.

[0076] Based on the aforementioned system architecture, unit dynamic model, and operational constraints, this invention employs a model predictive control algorithm to achieve closed-loop automatic power generation control of a photovoltaic-thermal multi-energy complementary system. Through rolling optimization and real-time correction control logic, it adapts to the system's time-varying characteristics and multiple constraints. The specific control steps are as follows: S1. Establish a dynamic model for automatic power generation control of a photovoltaic-thermal multi-energy complementary system, including a photovoltaic power plant, a solar thermal power plant, a thermal storage system, a system frequency response component, a tie-line power exchange component, and a model predictive controller.

[0077] First, by integrating the core units of the system with the grid interface, a unified dynamic model for automatic generation control across the entire domain is constructed. This model comprehensively covers the dynamic response unit of the photovoltaic power plant, the multi-stage power response unit of the solar thermal power plant, the energy iteration unit of the thermal storage system, the system frequency response unit, the tie-line power transmission unit, and the model predictive control decision unit. This model can characterize the power response timing characteristics, energy conversion laws, and grid parameter coupling relationships of each unit, providing model support for subsequent state prediction, deviation correction, and optimization solutions.

[0078] To uniformly describe the dynamic coupling relationship between photovoltaic power plants, solar thermal power plants, thermal storage systems, and grid interface links, the continuous dynamic model of the system is represented as follows: (43) The system output equation is: (44) In the formula, For system state variables, To control variables, For the disturbance variable, For output variables, A It is a continuous state matrix. B To control the input matrix, E The perturbation input matrix is... C This is the output matrix.

[0079] Among them, the system frequency response element is used to describe the frequency change process under the combined effects of photovoltaic power plants, solar thermal power plants, load disturbances, and tie-line power exchange deviations.

[0080] The system frequency dynamic equation and tie-line dynamic equation are as follows: (45) (46) in, H The system's equivalent inertial time constant. D This is the load damping coefficient. This is the synchronization coefficient for the tie line.

[0081] S2, collect system frequency deviation, tie line switching power deviation, current output power of photovoltaic power station, current output power of solar thermal power station, current heat storage of thermal storage system and load disturbance change, and construct regional control deviation.

[0082] At the start of each fixed control cycle, the system collects real-time operating status parameters across the entire domain, specifically including system frequency deviation, tie-line switching power deviation, current actual output power of the photovoltaic power plant, current actual output power of the solar thermal power plant, current remaining heat storage capacity of the thermal storage system, and real-time load disturbance changes. To facilitate the unified use of real-time measurement information by the model predictive controller, the [database name missing] can be [database name missing]. k The operating parameters collected in each control cycle are organized into a real-time measurement vector: (47) The system frequency deviation is determined by the difference between the measured system frequency and the rated frequency. (48) Based on the collected system frequency deviation and tie-line switching power deviation, a regional control deviation is constructed using a tie-line power-frequency deviation control method: (49) In the formula, For the first k The real-time measurement vectors acquired in each control cycle For the first k Real-time system measured frequency The system's rated frequency, For the first k Time zone control deviation The system frequency deviation coefficient, For the first k Constant-time communication line switching power deviation For the first k The current actual output power of the photovoltaic power station at any time. For the first k The current actual output power of the solar thermal power plant For the first k The current heat storage capacity of the thermal storage system is constantly monitored. For the first k The amount of change in load disturbance at any given time.

[0083] Through the above three equations, the model predictive controller can obtain the real-time operating status of the system in each control cycle and couple the frequency deviation and tie-line exchange power deviation into the area control deviation, providing real-time input for subsequent discrete state-space prediction, rolling optimization objective function construction and feedback correction.

[0084] S3. Based on the dynamic response relationship between photovoltaic power plants and solar thermal power plants, establish a discrete state-space prediction model.

[0085] By considering the dynamic response coupling relationships of photovoltaic power plants, solar thermal power plants, thermal storage systems, and the power grid, the continuous system model is discretized to establish a discrete state-space prediction model adapted to MPC rolling optimization, predicting the changing trends of various state variables of the system in the future time domain. The continuous state-space model is as follows: (50) Discretizing the continuous state-space model yields: (51) In the formula, , , , These are the discrete state matrix, control input matrix, disturbance matrix, and output matrix, respectively.

[0086] To clarify the model variable definitions and unify the system parameter dimensions, the definitions of state variables, control variables, disturbance variables, and output variables are as follows. State variables are: (52) The control variables are: (53) The disturbance variable is: (54) The output variable is: (55) in, This is an independently controllable quantity for photovoltaic power plants. This refers to the independent control variables of a solar thermal power plant. This model directly solves for the independent control variables of the two units, abandoning the traditional secondary allocation mode of the total control variable and avoiding control allocation deviations from the source.

[0087] S4. Set the prediction time domain N p and control time domain N c Construct a model predictive control rolling optimization objective function.

[0088] Based on the system's requirements for adjustment response speed and control accuracy, the prediction time domain is set. N p With control time domain N c The predicted future state sequence, future control sequence, and future output sequence in the time domain are as follows: (56) (57) (58) Meanwhile, the perturbation sequence in the prediction time domain is defined as: (59) In the formula, For the first k The predicted time-domain perturbation sequence constructed at each time point. For the first k Time for the first k + i The estimated value of the perturbation variable at any given time. This invention does not introduce external resource prediction data; the perturbation sequence does not include solar irradiance, wind speed, DNI, or predicted available output of new energy sources. In specific implementation, It can be determined based on the current load disturbance measurement value, the dispatch given disturbance value, or the short-term hold value, and is updated in the next control cycle through real-time measurement and feedback correction.

[0089] Based on the discrete state-space model and the definitions of the state sequence, control sequence, and disturbance sequence in the prediction time domain, the expression for the state sequence in the prediction time domain can be obtained: (60) The expression for the output sequence in the prediction time domain is: (61) In the formula, Φ, Γ, Γ w Ω are the prediction matrices constructed from the discrete state-space model.

[0090] S5. Under the constraints of dynamic reserve margin of photovoltaic power plant, output of solar thermal power plant, ramp-up rate of solar thermal unit, capacity of thermal storage system, active power balance of system, frequency security and alternating power deviation of tie line, solve for the optimal control sequence in the future control time domain.

[0091] To ensure that the optimization results closely reflect the actual engineering situation and meet the requirements for safe equipment operation, all system operating constraints are introduced during the optimization process. The optimal control sequences for photovoltaic and solar thermal power plants are then solved in a rolling manner within the feasible region, considering future control time domains. Specifically, the dynamic reserve margin constraint for the photovoltaic power plant is: (62) in, For the first k The dynamic reserve margin that can be increased by the photovoltaic power station at any time. For the first k The dynamic adjustment margin that can be used to adjust photovoltaic power plants at any time.

[0092] (63) (64) In the formula, The active power output of the photovoltaic power station at time k is... This represents the current maximum active power capacity allowed for automatic generation control in photovoltaic power plants. This represents the current minimum active power allowed for photovoltaic power plants to participate in automatic power generation control.

[0093] The output constraints of the solar thermal power plant are: (65) The ramp rate constraint for solar thermal power units is: (66) In the formula, and These represent the minimum and maximum active power output of a solar thermal power plant, respectively. and These are the maximum downhill ramp rate and maximum uphill ramp rate of the solar thermal power unit, respectively. The capacity of the thermal storage system and the terminal retention constraints are as follows: (67) (68) In the formula, and These represent the minimum and maximum heat storage capacity of the thermal storage system, respectively. To predict the lower limit of heat storage at the end of the time domain.

[0094] The system's active power balance constraint is expressed in incremental form as follows: (69) In the formula, For the increase in output power of photovoltaic power plants, For the increase in output power of solar thermal power plants, This represents the change in load disturbance. For the offset of the switching power of the tie line, This is a power balance slack variable.

[0095] The tie-line switching power deviation constraint is: (70) Frequency security constraints: (71) In the formula, , These are the lower and upper limits of the allowable deviation for the switching power of the tie line, respectively. , These represent the lower and upper limits of the allowable system frequency deviation, respectively.

[0096] Under the above constraints, the model predictive controller solves for the optimal control sequence in the future control time domain: (72) S6. Take the first control quantity in the optimal control sequence and apply it to the photovoltaic power station and the solar thermal power station.

[0097] Considering the dynamic changes in the real-time operating conditions of the power system, control sequences with excessively long time domains have no practical execution value. Therefore, this invention adopts a rolling execution mechanism, extracting only the first control quantity from the optimal control sequence as the execution instruction for the current cycle, and issuing it to the photovoltaic power station and the solar thermal power station respectively to complete the single-cycle active power regulation. (73) Right now: (74) In the formula, This represents the optimal control quantity issued to the photovoltaic power station during the current control cycle. This is the optimal control quantity issued to the solar thermal power plant for the current control cycle. Subsequently, The data will be sent to the active power control system of the photovoltaic power station. Issued to the AGC execution stage of the solar thermal power plant.

[0098] S7. In the next control cycle, the initial state of the prediction model is corrected based on the deviation between the actual output of the system and the predicted output of the previous control cycle, and S2 to S6 are repeated to realize the closed-loop automatic power generation control of the photovoltaic and solar thermal multi-energy complementary system.

[0099] To eliminate prediction biases caused by uncertainties such as model mismatch, random load disturbances, and output fluctuations, and to ensure the long-term prediction accuracy of the model, this system incorporates a feedback correction mechanism. After a single-cycle control command is executed, the actual output state variables of the system are collected and compared with the predicted output values ​​of the previous cycle. The prediction error is calculated, and the initial prediction values ​​for the next cycle are corrected, thus achieving closed-loop iterative optimization. The prediction error is: (75) The predicted initial value for the next control cycle after feedback correction is: (76) In the formula, For prediction error, This is the actual measured output. To predict the output, The feedback correction gain matrix is ​​used. After model correction is completed, the system returns to step S2 to start the state acquisition, optimization solution and command execution of the next control cycle, and repeats the cycle to realize continuous, stable and high-precision automatic power generation closed-loop control of the photovoltaic and solar thermal multi-energy complementary system.

[0100] To comprehensively verify the effectiveness and engineering applicability of the control method proposed in this invention, this embodiment constructs a power system simulation model adapted to the photovoltaic-thermal multi-energy complementary scenario, reproducing typical operating conditions commonly encountered in actual power grid operation, such as load step disturbances, continuous load fluctuations, and sudden drops in photovoltaic output. Simultaneously, a PID feedback distribution control method and a conventional MPC control method are set as control groups, and comparative tests are conducted on aspects such as frequency response characteristics, smoothness of photovoltaic-thermal output, thermal storage state maintenance capability, and control command executability. The corresponding simulation results are compared as follows: Figure 6 As shown.

[0101] The simulation uses a discrete-time approach with a sampling period of 1 second, and the model predictive controller predicts in the time domain. N p Set to 30 to control the time domain. N c Assuming a value of 30, the system rated frequency is 50Hz, and the area control deviation is constructed using a tie-line power-frequency deviation control method. The rated capacity of the solar thermal power plant is set to 100MW, with an AGC time delay constant of 20s, a governor time constant of 0.1s, a turbine reheat coefficient of 2.5, a reheat time constant of 10s, a generator time constant of 0.2s, a load damping coefficient of 2, a unit moment of inertia of 10s, a unit time constant of 10s, and a unit droop coefficient of 0.1. The rated capacity of the photovoltaic power plant is set to 50MW, with a unit time constant of 2s, an AGC time delay constant of 1s, an inverter time constant of 0.01s, and a unit droop coefficient of 0.02.

[0102] In the continuous load disturbance simulation, the initial system state was that the concentrated solar power (CSP) plant was operating stably at 60MW, with a system frequency of 50Hz. Subsequently, load disturbances were applied at multiple times, with the disturbance sequence as follows: load increased to 72MW at 60s, decreased to 65MW at 150s, decreased to 58MW at 260s, increased to 65MW at 320s, increased to 72MW at 400s, and returned to 60MW at 460s. Simulation results showed that when the load suddenly increased from 60MW to 72MW, the system frequency decreased. The model predictive controller resolved the control sequence based on the frequency deviation and system state, and issued an additional power output command to the CSP plant. The CSP plant output increased to approximately 71MW within about 90s, at which point the system frequency recovered to 50.007Hz, with a deviation from the rated frequency of 50Hz not exceeding 0.01Hz. When the load dropped to 65MW, the system frequency briefly deviated to 50.039Hz. The model predictive controller corrected the control input through feedback correction and rolling optimization, reducing the output of the concentrated solar power (CSP) plant to 64.66MW within approximately 110 seconds, and restoring the frequency to 49.997Hz. When the load further dropped to 58MW, the CSP plant reduced its output to 58MW within approximately 60 seconds, with the system frequency at 49.999Hz, which is close to the rated frequency. This demonstrates that the AGC controller of the CSP plant based on model predictive control can rapidly adjust the output and smooth frequency fluctuations under multiple consecutive load disturbances.

[0103] In the simulation of photovoltaic (PV) and solar thermal power (CSP) coordinated control under a sudden drop in PV output, both the PV and CSP plants are initially in stable operation. During the simulation, a 5MW active power output drop occurs at 120 seconds in the PV plant to simulate short-term output reduction caused by cloud cover, power limitations within the plant, or inverter control constraints. This embodiment does not incorporate external illumination or output prediction data; the PV output drop is only considered a real-time power disturbance measured by the system and enters the model predictive controller. After the PV plant's output drops, the system experiences an active power deficit, and the system frequency decreases accordingly. The model predictive controller receives new system frequency deviations, tie-line switching power deviations, real-time PV and CSP outputs, and the amount of heat stored in the thermal storage system. It then reconstructs the state variables and performs rolling optimization. Because the PV plant is connected to the grid via power electronic devices, its response speed is relatively fast. Therefore, within its dynamic reserve margin, it prioritizes rapid power compensation during the initial stage of the disturbance. The CSP plant, constrained by ramp-up rate and thermal storage status, gradually increases its output power to undertake subsequent continuous support tasks. This simulation condition corresponds to the photovoltaic dynamic reserve margin constraint, solar thermal ramp-up constraint, and thermal storage state constraint in this embodiment, and does not use external resource prediction conditions.

[0104] To further verify the control effect of the method described in this invention, this embodiment sets up three control methods for comparison. The first type is the PID feedback allocation control method, which generates system adjustment commands based on regional control deviations and then allocates control quantities according to the adjustment capabilities of photovoltaic power plants and solar thermal power plants. The second type is the ordinary MPC control method, which directly solves the control quantities using a state-space prediction model and rolling optimization mechanism, but does not introduce end-point retention constraints of the thermal storage system and dynamic reserve margin constraints of the photovoltaic power plant. The third type is the MPC control method of this invention, which directly solves the control quantities of the photovoltaic power plant and solar thermal power plant in the prediction time domain, and simultaneously considers the dynamic reserve margin constraints of the photovoltaic power plant, the output constraints of the solar thermal power plant, the ramp-up rate constraints of the solar thermal unit, the capacity constraints of the thermal storage system, the end-point thermal storage retention constraints, the frequency safety constraints, and the tie-line switching power deviation constraints.

[0105] Reference Figure 6 Simulation results show that, compared with PID feedback distribution control and ordinary MPC control, the control method described in this invention has better control performance in terms of frequency recovery speed, overshoot suppression, and solar thermal power output smoothness. PID control mainly relies on current deviation feedback and control quantity distribution, resulting in a relatively slow frequency recovery process and significant overshoot in solar thermal power plant output. Ordinary MPC control can shorten the frequency recovery time through rolling optimization, but due to insufficient consideration of photovoltaic dynamic reserve margin and thermal storage end-point maintenance constraints, certain control quantity abrupt changes and rapid thermal storage consumption may still occur. The MPC control of this invention simultaneously considers the photovoltaic rapid response capability, solar thermal ramp-up rate, and thermal storage state constraints. It can quickly compensate by calling upon the photovoltaic power plant's dynamic reserve in the early stages of disturbances, and the solar thermal power plant will provide subsequent continuous support, thereby achieving rapid frequency recovery, smooth power output adjustment, and maintenance of thermal storage state.

[0106] In summary, the simulation results of this embodiment demonstrate that the present invention can achieve automatic power generation control of a photovoltaic-thermal multi-energy complementary system without introducing external resource prediction data, relying on real-time state acquisition, discrete state-space prediction models, rolling optimization, and feedback correction. This method can effectively reduce system frequency deviation and power overshoot, shorten frequency recovery time, prevent excessive energy release by the thermal storage system during short-term frequency regulation, and improve the control accuracy, operational stability, and engineering applicability of the photovoltaic-thermal multi-energy complementary system.

Claims

1. An automatic power generation control method for a photovoltaic-thermal multi-energy complementary system, characterized in that, Includes the following steps: S1. Establish an automatic power generation control dynamic model for a photovoltaic-thermal multi-energy complementary system, including a photovoltaic power plant, a solar thermal power plant, a thermal storage system, a system frequency response component, a tie-line power exchange component, and a model predictive controller. S2, collect system frequency deviation, tie line switching power deviation, current output power of photovoltaic power station, current output power of solar thermal power station, current heat storage of thermal storage system and load disturbance change, and construct regional control deviation; S3. Based on the dynamic response relationship between photovoltaic power plants and solar thermal power plants, establish a discrete state-space prediction model; ； in, For system state variables, To control variables, For the disturbance variable, For output variables, , , , These are the discrete state matrix, control input matrix, disturbance matrix, and output matrix, respectively. S4. Set the prediction time domain N p and control time domain N c Construct the model predictive control rolling optimization objective function: ； in ; S5. Under the constraints of dynamic reserve margin of photovoltaic power plant, output of solar thermal power plant, ramp-up rate of solar thermal unit, capacity of thermal storage system, active power balance of system, frequency security and power deviation of tie line, solve the optimal control sequence in the future control time domain. S6. Take the first control quantity in the optimal control sequence and apply it to the photovoltaic power station and the solar thermal power station; S7. In the next control cycle, the initial state of the prediction model is corrected by feedback based on the deviation between the actual output of the system and the predicted output of the previous control cycle, and S2 to S6 are repeated to realize the closed-loop automatic power generation control of the photovoltaic and solar thermal multi-energy complementary system.

2. The method according to claim 1, characterized in that: The regional control deviation is constructed using a tie-line power-frequency deviation control method: ； In the formula, This is the frequency deviation coefficient. This is the system frequency deviation. This refers to the power deviation of the tie line.

3. The method according to claim 2, characterized in that: In step S3, the system state variable is: ； In the formula, For the increase in output power of photovoltaic power plants, For the increase in output power of solar thermal power plants, For storing heat in the thermal storage system, For the valve position increment of the solar thermal unit speed controller; The control variable is: ； in, The control quantity is issued to the photovoltaic power station. This is the control quantity issued to the solar thermal power plant.

4. The method according to claim 1, characterized in that: The dynamic response model of the photovoltaic power station is as follows: ； in, The delay constant of photovoltaic AGC. The inverter time constant, The photovoltaic power response time constant is For photovoltaic power response gain; The dynamic reserve margin constraint of the photovoltaic power station is: ； in, For the first k The dynamic reserve margin that can be increased by the photovoltaic power station at any time. For the first k The dynamic adjustment margin that can be used to adjust photovoltaic power plants at any time.

5. The method according to claim 4, characterized in that: The adjustable dynamic reserve margin and adjustable dynamic adjustment margin of the photovoltaic power station are as follows: ； In the formula, The active power output of the photovoltaic power station at time k is... This represents the current maximum active power capacity allowed for automatic generation control in photovoltaic power plants. This represents the current minimum active power allowed for photovoltaic power plants to participate in automatic power generation control.

6. The method according to claim 1, characterized in that: The active power response process of a solar thermal power plant includes AGC delay, governor, reheat turbine, and generator power response stages; the control command after the AGC delay is: ； In the formula, s For the Laplace operator, The delay constant for AGC commands in a solar thermal power plant. The controller sends control signals to the solar thermal power plant. The control command is the result of the AGC delay period. The dynamic relationship of the speed governor can be expressed as: ； In the formula, For the incremental position of the governor valve of the solar thermal unit. For speed controller gain, The time constant of the speed controller, R This is the unit droop coefficient. This refers to the system frequency deviation. The mechanical power output of the reheat turbine stage is: ； In the formula, For the increase in mechanical power of the steam turbine, For turbine power gain, The time constant of the steam turbine. The reheat coefficient, The reheat time constant; The output power increment of the solar thermal power plant meets the following requirements: ； In the formula, For the increase in output power of solar thermal power plants, Let be the power response time constant of the solar thermal generator. Based on the above dynamic response model of the solar thermal power plant; The output constraint of the solar thermal power plant is: ； The ramp rate constraint for the solar thermal unit is: ； In the formula, and These represent the minimum and maximum active power output of a solar thermal power plant, respectively. and These are the maximum downhill and uphill speeds of the solar thermal power unit, respectively.

7. The method according to claim 1, characterized in that: The energy state of the thermal storage system satisfies: ； In the formula, For the first k The equivalent heat charging power that constantly enters the thermal storage system. To supply thermal power to the solar thermal power generation modules of the thermal storage system, For thermoelectric conversion efficiency, This is the heat storage loss coefficient.

8. The method according to claim 7, characterized in that: The capacity constraints of the thermal storage system include upper and lower limits for the amount of heat storage and constraints for maintaining thermal storage at the terminal: ； In the formula, and These represent the minimum and maximum heat storage capacity of the thermal storage system, respectively. To predict the lower limit of heat storage at the end of the time domain.

9. The method according to claim 1, characterized in that: The system frequency dynamic equation and tie-line dynamic equation are as follows: ； in, H The system's equivalent inertial time constant. D This is the load damping coefficient. This is the synchronization coefficient for the tie line; The active power balance constraint of the system adopts an incremental form: ； In the formula, This represents the change in load disturbance. For power balance slack variables, For the increase in output power of photovoltaic power plants, For the increase in output power of solar thermal power plants, This is to compensate for the power deviation of the tie line.

10. The method according to claim 1, characterized in that: The feedback correction includes: ； In the formula, For prediction error, For the first k Actual measured output at time +1 For the first k Time for the first k Predicted output at time +1 For feedback correction gain matrix, This is the initial value predicted for the next control cycle after feedback correction.

Images