Excitation and static frequency converter collaborative optimization control system and method for pumping start of pumped storage unit
By using real-time data acquisition and collaborative optimization control, the problems of response lag and poor coordination between the excitation system and the static frequency converter were solved, achieving optimal dynamic performance and improved startup efficiency during the pumped storage unit startup process, thus enhancing the smoothness and reliability of startup.
Patent Information
- Application Number
- CN202610012027.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-06
- Publication Date
- 2026-05-19
AI Technical Summary
When traditional pumped storage units start up, the excitation system and the control of the static frequency converter suffer from problems such as response lag, torque pulsation, poor coordination, and non-optimal start-up efficiency, resulting in an unsmooth start-up process, poor reliability, and high energy consumption.
The system uses a data acquisition and status awareness module to collect system status data in real time. The collaborative optimization controller performs rolling optimization based on a joint prediction model to calculate the optimal control command sequence, which is then synchronously sent to the excitation system and static frequency converter through a high-speed communication network to achieve deep collaboration and forward-looking control.
The system achieves optimal dynamic performance during the startup process of pumped storage units, reduces torque pulsation, improves startup smoothness and reliability, enhances startup efficiency and anti-interference capabilities, and realizes an intelligent startup strategy.
Smart Images
Figure CN122068558A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of pumped storage unit control technology, and in particular to a coordinated optimization control system and method for excitation and static frequency converters for pumped start-up of pumped storage units. Background Technology
[0002] When a pumped-storage unit starts up in pumping mode, it needs to be pulled from a standstill to synchronous speed using a static frequency converter (SFC). The traditional control method is a master-slave mode of "SFC master control, excitation follow". That is, the SFC issues frequency commands and excitation current reference values according to a preset speed curve, and the excitation system passively follows.
[0003] This traditional method has the following drawbacks: 1. Response lag and torque pulsation: The excitation system only responds passively and cannot predict the control intention of the SFC. The delay in the control cycle of the two leads to inaccurate torque coordination during the dynamic process, resulting in torque pulsation and increasing mechanical stress.
[0004] 2. Poor coordination: The two systems are controlled independently, lacking in-depth information interaction and global optimization objectives. When faced with sudden load changes or power grid disturbances, their anti-interference ability is weak, which can easily lead to startup failure.
[0005] 3. Non-optimal startup efficiency: To ensure reliability, a conservative startup curve (such as a fixed speed ramp rate and excitation current) is usually adopted, which fails to achieve the fastest or most energy-efficient startup based on the real-time status of the unit (such as water temperature, bearing temperature, rotor heat capacity). Summary of the Invention
[0006] The purpose of this invention is to overcome the shortcomings of the prior art and provide a coordinated optimization control system and method for excitation and static frequency converters (SFCs) for pumping start-up of pumped storage units, so as to achieve deep coordination and forward-looking control between the excitation system and SFCs, and improve the smoothness, speed, reliability and energy saving of the pumping start-up process of pumped storage units.
[0007] To achieve the above-mentioned technical features, the objective of this invention is as follows: a coordinated optimization control system for excitation and static frequency converters for pumped-storage unit pumping start-up, the system comprising: The data acquisition and status awareness module is used to collect system status data in real time. The system status data includes unit speed, rotor position, stator current, stator voltage, rotor current, rotor voltage, static frequency converter (SFC) output frequency, SFC output voltage, excitation current, temperature of key parts of the unit, and instructions from the power grid dispatch system. A collaborative optimization controller, connected to the SFC and excitation system via a high-speed communication bus, internally includes a joint prediction model, a rolling optimizer, and a constraint processing unit. The joint prediction model is an integrated dynamic mathematical model encompassing the SFC, excitation system, synchronous motor, and pump load, used to predict the system's behavior over a future period. The rolling optimizer, in each control cycle, aims to improve startup performance by solving a constrained optimization problem based on the current system state and the joint prediction model, calculating the optimal control command sequence for the SFC and excitation system in the future control time domain. The constraint processing unit ensures that the optimization calculations always meet the system's safe operation constraints. A high-speed execution and communication network is used to send the optimal control commands calculated by the collaborative optimization controller to the SFC and excitation system in real time and synchronously.
[0008] Preferably, the key components of the unit include rotor windings and bearings.
[0009] Preferably, the startup performance includes shortening startup time, reducing torque pulsation, and reducing losses.
[0010] Preferably, the safety operation constraints include the maximum SFC current, the excitation system peak voltage, the rotor winding temperature rise limit, and the maximum vibration value of the unit.
[0011] Preferably, the collaborative optimization controller uses a high-performance multi-core industrial computer or a high-end PLC with floating-point computing capabilities.
[0012] Preferably, the high-speed execution and communication network adopts the real-time Ethernet protocol, which is either the EtherCAT protocol or the PROFINETIRT protocol.
[0013] Preferably, the data acquisition and status sensing module accesses sensor data from the unit site via real-time Ethernet or a high-speed I / O module. The sensor data includes unit speed and rotor position signals from the encoder, terminal voltage, stator current, and rotor current signals from the voltage / current transformer, and rotor winding, bearing, and cooling water temperature signals from the temperature sensor.
[0014] Another aspect of the present invention provides a method for coordinated optimization control of excitation and static frequency converters for pumped start-up of pumped storage units, the method being implemented based on any one of the systems described above, comprising: Step S1: Receive the pumping start command, initialize the system, and load the current unit status parameters and head information; Step S2: The collaborative optimization controller performs rolling optimization based on the joint prediction model; Step S3: In the next control cycle, repeat step S2 to achieve closed-loop optimization control of the entire startup process in a rolling manner.
[0015] Step S4: When the unit speed reaches the synchronous speed, the startup process ends and the system switches to normal operating mode.
[0016] Preferably, step S2 specifically includes: S2.1, Status Acquisition: Acquire real-time data from the data acquisition and status awareness module; S2.2, Model Prediction: Using a joint prediction model, predict the system's state trajectory over a future period under different SFC output voltage / frequency commands and excitation voltage commands. The state trajectory includes speed, torque, power angle, current, and temperature. S2.3, Online Optimization: Solve the optimization problem to minimize the performance index function J, where the general formula for the performance index function J is: ; S2.4, Instruction Issuance: The first element of the control instruction sequence obtained from the optimized calculation, i.e. the optimal instruction at the current moment, is synchronously sent to the SFC and the excitation system. The instruction sent to the SFC includes the output voltage amplitude, frequency, and phase, and the instruction sent to the excitation system includes the excitation voltage or current setpoint.
[0017] Preferably, in the online optimization process of step S2, the interior point method or the effective set method is used to solve the optimization problem, and the numerical computing library used is qpOASES or OSQP.
[0018] Preferably, the model prediction process in step S2 adopts the linear time-varying MPC method, which linearizes the joint prediction model around the current operating point in each sampling period to obtain a linear state-space model (A,B,C,D).
[0019] Preferably, in step S3, during the repetition of step S2, a state estimation step is also included, in which a Kalman filter is used to estimate the state to eliminate noise and to observe the load torque T_load in real time.
[0020] Preferably, in step S4, when the unit speed reaches 95%-100% of the synchronous speed and the power angle is stable, the system issues a "synchronous grid connection" command, the SFC exits, and the excitation system switches to the normal voltage regulation mode.
[0021] The present invention has the following beneficial effects: The collaborative optimization control system and method of this invention collects system state data in real time through a data acquisition and state perception module. The collaborative optimization controller performs rolling optimization based on a joint prediction model, solving the constrained optimization problem in each control cycle to obtain the optimal control command sequence. This sequence is then synchronously sent to the SFC and excitation system via a high-speed execution and communication network, achieving deep collaboration and forward-looking control between the excitation system and SFC. Compared with the prior art, this invention has the following advantages: 1. Global Dynamic Optimization: By using the Model Predictive Control (MPC) framework, SFC and excitation are treated as a whole for collaborative optimization, achieving optimal dynamic performance throughout the startup process, rather than local or instantaneous optimization.
[0022] 2. Foresight and smoothness: MPC's "predictive-optimization" mechanism enables the control system to anticipate future dynamics and adjust SFC and excitation commands in advance, fundamentally eliminating response lag and achieving extremely smooth torque control, effectively reducing mechanical stress.
[0023] 3. Strong robustness: Through rolling optimization and real-time feedback, the system can automatically compensate for model errors and external disturbances (such as load torque fluctuations and small changes in grid voltage), which greatly improves the success rate and reliability of startup.
[0024] 4. Adaptive and intelligent: The system can automatically generate the optimal start-up curve according to different water head and unit temperature conditions, realizing an intelligent start-up strategy of "fast start-up in cold state and stable start-up in hot state". Attached Figure Description
[0025] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0026] Figure 1 The structural block diagram of the collaborative optimization control system described in this invention.
[0027] Figure 2 A schematic diagram comparing the torque response of the traditional control method and the control method of the present invention during the startup process.
[0028] Figure 3 A flowchart of the collaborative optimization control method described in this invention. Detailed Implementation
[0029] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. For some well-known technologies, those skilled in the art may also use other alternative methods to implement the invention. Moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.
[0030] Example 1: like Figure 1 This is a structural block diagram of the collaborative optimization control system described in this invention. The "collaborative optimization controller" introduced in this invention is the core of the entire system. It collects status information from the generator set and external scheduling commands through a high-speed communication network, performs internal calculations, and then outputs globally optimized collaborative control commands to the SFC and excitation system, thereby forming a closed-loop intelligent control system.
[0031] A coordinated optimization control system for excitation and static frequency converters for pumped start-up of pumped storage units, the system comprising: It includes a data acquisition and status awareness module, a collaborative optimization controller, and a high-speed execution and communication network.
[0032] The data acquisition and status awareness module is used to collect system status data in real time. This system status data includes unit speed, rotor position, stator current, stator voltage, rotor current, rotor voltage, SFC output frequency, SFC output voltage, excitation current, rotor winding temperature, bearing temperature, and commands from the power grid dispatching system. The data acquisition and status awareness module accesses sensor data from the unit site via a high-speed I / O module, including unit speed and rotor position signals from the encoder, terminal voltage, stator current, and rotor current signals from the voltage / current transformer, and rotor winding, bearing, and cooling water temperature signals from temperature sensors.
[0033] The collaborative optimization controller employs a high-end PLC based on the ARM Cortex-A series architecture, connected to the SFC and excitation system via a high-speed communication bus. Internally, the controller includes a joint prediction model, a rolling optimizer, and a constraint processing unit. The joint prediction model is an integrated dynamic mathematical model encompassing the SFC, excitation system, synchronous motor, and pump load, used to predict the system's behavior over a future period. The rolling optimizer, in each control cycle, aims to shorten start-up time, reduce torque ripple, and lower losses. Based on the current system state and the joint prediction model, it solves a constrained optimization problem, calculating the optimal control command sequence for the SFC and excitation system in the future control time domain. The constraint processing unit ensures that the optimization calculations always meet the system's safe operating constraints, including a maximum SFC current of 1500A, a maximum excitation system voltage of 800V, a rotor winding temperature rise limit of 80K, and a maximum unit vibration of 0.05mm.
[0034] The high-speed execution and communication network uses the EtherCAT protocol to transmit the optimal control commands calculated by the collaborative optimization controller to the SFC7 and excitation system in real time and synchronously. The collaborative optimization controller communicates with the power plant monitoring system via the standard TCP / IP protocol, receiving start / stop commands and head signals, and uploading panoramic data of the start-up process.
[0035] The collaborative optimization control system in this embodiment collects comprehensive system status data in real time through the data acquisition and status perception module. The collaborative optimization controller, based on the ARM Cortex-A series PLC, has powerful computing capabilities and can quickly perform rolling optimization calculations. The EtherCAT protocol ensures the real-time synchronous issuance of control commands, realizing deep collaborative control between the excitation system and SFC, and improving the performance of the pumped storage unit's pumping start-up process.
[0036] Example 2: See Figure 3 The flowchart of the collaborative optimization control method described in this invention. The control method of this invention is a cyclically executed rolling optimization process. In each control cycle, it re-predicts and optimizes the control strategy for a future period based on the latest system state, and only executes the first step. This "rolling optimization, feedback correction" mechanism enables the control system to continuously adapt to system changes, ensuring that the startup process always follows the optimal or suboptimal trajectory. This invention provides a collaborative optimization control method for excitation and static frequency converters for pumped-storage unit pumping startup, comprising: Step S1: Receive the pumping start command, initialize the system, and load the current unit status parameters and head information; Step S2: The collaborative optimization controller performs rolling optimization based on the joint prediction model; S2.1, Status Acquisition: Acquire real-time data from the data acquisition and status awareness module; S2.2, Model Prediction: Using a joint prediction model, predict the system's state trajectory over a future period under different SFC output voltage / frequency commands and excitation voltage commands. The state trajectory includes speed, torque, power angle, current, and temperature. S2.3, Online Optimization: Solve the optimization problem to minimize the performance index function J, where the general formula for the performance index function J is: ; S2.4, Instruction Issuance: The first element of the control instruction sequence obtained from the optimized calculation, i.e. the optimal instruction at the current moment, is synchronously sent to the SFC and the excitation system. The instruction sent to the SFC includes the output voltage amplitude, frequency, and phase, and the instruction sent to the excitation system includes the excitation voltage or current setpoint.
[0037] Step S3: In the next control cycle, repeat step S2 to achieve closed-loop optimization control of the entire startup process in a rolling manner.
[0038] Step S4: When the unit speed reaches the synchronous speed, the startup process ends and the system switches to normal operating mode.
[0039] Example 3: 1. System hardware configuration and interfaces: The core hardware of this invention's system is a collaborative optimization controller, which can be a high-performance multi-core industrial computer or a high-end PLC with powerful floating-point computing capabilities (such as a controller based on the ARM Cortex-A series or Intel x86 architecture). Its interface with existing systems is as follows: Interface with the SFC and excitation system: Real-time Ethernet protocols such as EtherCAT or PROFINETIRT are used. These two protocols offer microsecond-level synchronization accuracy and extremely low communication jitter, sufficient to meet the high requirements of control cycle synchronization in this invention. The collaborative optimization controller acts as the master station, and the controllers of the SFC and excitation system act as slave stations.
[0040] Interface with the data acquisition module: Through the aforementioned real-time Ethernet or high-speed I / O module, sensor data from the unit's field is received, including: The unit speed and rotor position signals are from the encoder.
[0041] Terminal voltage, stator current, and rotor current from voltage / current transformers (converted by transmitters).
[0042] Temperature signals from the rotor windings (via wireless temperature measurement or slip ring transmission), bearings, and cooling water are received from temperature sensors.
[0043] Interface with human-machine interface and upper-level system: Communicates with the power plant monitoring system through standard TCP / IP protocol, receives start / stop commands and head signals, and uploads panoramic data of the start-up process.
[0044] 2. Establishment of the joint prediction model: The key to this invention lies in a sufficiently accurate and computationally efficient joint prediction model. This model is a discrete-time state-space model, and its core state variables, control variables, and model equations are as follows: State variable (x[k]): ω_r[k]: Mechanical angular velocity of the unit rotor; δ[k]: Rotor power angle (phase relative to SFC output voltage); Ψ_fd[k]: Rotor excitation flux linkage; I_sd[k], I_sq[k]: Direct and quadrature axis components of stator current in the dq rotating coordinate system; T_load_hat[k]: Observed value of load torque (as an extended state); Control variable (u[k]): U_sfc[k]: The amplitude command for the SFC output voltage; f_sfc[k]: SFC output frequency command (corresponding to electric angular velocity ω_e=2πf_sfc); U_f[k]: Excitation voltage command output by the excitation system; Output variable (y[k]): T_e[k]: Electromagnetic torque of the generator; I_s[k]: Stator current amplitude (used for overcurrent protection constraints); V_term[k]: Terminal voltage; Model equations (discrete nonlinear form): This model is based on a practical fifth-order or third-order model of a synchronous motor in the dq coordinate system, and is simplified by incorporating the inverter characteristics of SFC.
[0045] Nonlinear state update equation: x[k+1]=f(x[k],u[k])+w[k]; Output equation: y[k]=g(x[k],u[k])+v[k]; in: f() and g() are nonlinear functions that contain the electrical and mechanical dynamic equations of the motor.
[0046] Electromagnetic torque: T_e=(3 / 2)*p*(Ψ_d*I_q-Ψ_q*I_d); Where p is the pole pair number, and Ψ_d and Ψ_q are the magnetic flux linkages along the dq axis.
[0047] w[k] and v[k] are process noise and measurement noise, respectively.
[0048] Model linearization: To reduce the computational load on the online platform, models f and g can be linearized around the current operating point (x0, u0) in each sampling period to obtain a linear state-space model (A, B, C, D), which is used for MPC prediction. This method, known as "linear time-varying MPC," is widely used in industry.
[0049] 3. Specific implementation of the rolling optimizer: Mathematical description of the optimization problem: In each control period k, solve the following constrained quadratic programming problem; minJ=Σ_{i=0}^{Np-1}[(T_e[k+i|k]-T_ref)^2*Q_T+ (ω_r[k+i|k]-ω_ref[k+i])^2*Q_ω+ ΔU_sfc[k+i|k]^2*R_sfc+ ΔU_f[k+i|k]^2*R_f] Constraints: State and output constraints: Stator current constraint: I_s_min≤I_s[k+i|k]≤I_s_max; Terminal voltage lower limit constraint: V_term_min≤V_term[k+i|k]; Torque limit (mechanical strength limit): T_e[k+i|k]≤T_max; Control variable constraints: U_sfc_min≤U_sfc[k+i|k]≤U_sfc_max; f_sfc_min≤f_sfc[k+i|k]≤f_sfc_max; U_f_min≤U_f[k+i|k]≤U_f_max; in: Np is the prediction time domain.
[0050] Q_T, Q_ω, R_sfc, and R_f are positive definite weight matrices (scalars).
[0051] T_ref is the desired electromagnetic torque, which can be set to a constant value slightly larger than the estimated load torque during the initial startup phase to ensure smooth acceleration.
[0052] ω_ref[k+i] is the desired rotational speed trajectory, which can be a smooth S-shaped curve.
[0053] [k+i|k] represents the predicted value at time k for time k+i.
[0054] Solution algorithm: Since the problem is transformed into a quadratic programming problem, an efficient interior-point method or effective set method is used for solving it. Mature numerical computing libraries such as qpOASES (a QP solver specifically for embedded systems) or OSQP are employed.
[0055] 4. Detailed description of the workflow (e.g.) Figure 3 ): (1) Initialization: The system is powered on and loads motor parameters, model parameters, weight coefficients and constraint limits. It receives the start command and head information.
[0056] (2) First period calculation: Collect the initial state x[0].
[0057] The QP problem is initialized based on the x[0] linearization model.
[0058] Solve QP to obtain the first control instruction u[0]=[U_sfc[0],f_sfc[0],U_f[0]].
[0059] U_sfc[0] and f_sfc[0] are sent to the SFC via the EtherCAT network, and U_f[0] is sent to the excitation system.
[0060] (3) Scrolling execution: Wait for the next control cycle (the cycle is usually set to 1-5ms).
[0061] Collect the new system state x[k].
[0062] State estimation: The state is estimated using a Kalman filter to eliminate noise and to monitor the load torque T_load in real time.
[0063] Model update: Update the linearized model (A,B,C,D) based on the latest x[k].
[0064] Solve QP: Solve the updated optimization problem to obtain the new control instruction sequence u[k].
[0065] Instruction issuance: The first instruction u[k] in the sequence is issued to the execution mechanism.
[0066] (4) Looping and Termination: Repeat step 3 until the unit speed reaches 95%-100% of the synchronous speed and the power angle is stable. Then, the system issues a "synchronous grid connection" command, the SFC exits, the excitation system switches to normal voltage regulation mode, and the startup process ends.
[0067] See Figure 2 A schematic diagram comparing the torque response during startup using traditional control methods and the control method of this invention. The diagram qualitatively compares the electromagnetic torque curves under the two methods. The traditional "master-slave" control method, due to response lag and poor coordination, results in significant torque pulsations during startup. In contrast, this invention, through forward-looking optimization control using MPC, can output extremely smooth torque, significantly reducing the impact on the mechanical system. (Qualitative only, not quantitative).
[0068] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.
Claims
1. A coordinated optimization control system for excitation and static frequency converters for pumped-storage unit pumping start-up, characterized in that, The system includes: The data acquisition and status awareness module is used to collect system status data in real time. The system status data includes unit speed, rotor position, stator current, stator voltage, rotor current, rotor voltage, static frequency converter (SFC) output frequency, SFC output voltage, excitation current, temperature of key parts of the unit, and instructions from the power grid dispatch system. A collaborative optimization controller, connected to the SFC and excitation system via a high-speed communication bus, internally includes a joint prediction model, a rolling optimizer, and a constraint processing unit. The joint prediction model is an integrated dynamic mathematical model encompassing the SFC, excitation system, synchronous motor, and pump load, used to predict the system's behavior over a future period. The rolling optimizer, in each control cycle, aims to improve startup performance by solving a constrained optimization problem based on the current system state and the joint prediction model, calculating the optimal control command sequence for the SFC and excitation system in the future control time domain. The constraint processing unit ensures that the optimization calculations always meet the system's safe operation constraints. A high-speed execution and communication network is used to send the optimal control commands calculated by the collaborative optimization controller to the SFC and excitation system in real time and synchronously.
2. The excitation and static frequency converter coordinated optimization control system for pumped-storage unit pumping start-up as described in claim 1, characterized in that, The key components of the unit include the rotor windings and bearings.
3. The excitation and static frequency converter coordinated optimization control system for pumped start-up of a pumped storage unit according to claim 1, characterized in that, The startup performance includes shortening startup time, reducing torque pulsation, and reducing losses.
4. The excitation and static frequency converter coordinated optimization control system for pumped start-up of a pumped storage unit according to claim 1, characterized in that, The safety operation constraints include the maximum SFC current, the peak voltage of the excitation system, the rotor winding temperature rise limit, and the maximum vibration value of the unit.
5. The excitation and static frequency converter coordinated optimization control system for pumped start-up of a pumped storage unit according to claim 1, characterized in that, The collaborative optimization controller uses a high-performance multi-core industrial computer or a high-end PLC with floating-point operation capabilities.
6. The excitation and static frequency converter coordinated optimization control system for pumped start-up of a pumped storage unit according to claim 1, characterized in that, The high-speed execution and communication network adopts the real-time Ethernet protocol, which is either the EtherCAT protocol or the PROFINETIRT protocol.
7. The excitation and static frequency converter coordinated optimization control system for pumped start-up of a pumped storage unit according to claim 1, characterized in that, The data acquisition and status sensing module accesses sensor data from the unit site via real-time Ethernet or high-speed I / O module. The sensor data includes unit speed and rotor position signals from the encoder, terminal voltage, stator current, and rotor current signals from the voltage / current transformer, and rotor winding, bearing, and cooling water temperature signals from the temperature sensor.
8. A method for coordinated optimization control of excitation and static frequency converters for pumped start-up of pumped storage units, characterized in that, The method is implemented based on the system described in any one of claims 1-7, and includes: Step S1: Receive the pumping start command, initialize the system, and load the current unit status parameters and head information; Step S2: The collaborative optimization controller performs rolling optimization based on the joint prediction model; Step S3: In the next control cycle, repeat step S2 to achieve closed-loop optimization control of the entire startup process in a rolling manner. Step S4: When the unit speed reaches the synchronous speed, the startup process ends and the system switches to normal operating mode.
9. A method for coordinated optimization control of excitation and static frequency converters for pumped start-up of a pumped storage unit according to claim 8, characterized in that, Step S2 specifically includes: S2.1, Status Acquisition: Acquire real-time data from the data acquisition and status awareness module; S2.2, Model Prediction: Using a joint prediction model, predict the system's state trajectory over a future period under different SFC output voltage / frequency commands and excitation voltage commands. The state trajectory includes speed, torque, power angle, current, and temperature. S2.3, Online Optimization: Solve the optimization problem to minimize the performance index function J, where the general formula for the performance index function J is: ； S2.4, Instruction Issuance: The first element of the control instruction sequence obtained from the optimized calculation, i.e. the optimal instruction at the current moment, is synchronously sent to the SFC and the excitation system. The instruction sent to the SFC includes the output voltage amplitude, frequency, and phase, and the instruction sent to the excitation system includes the excitation voltage or current setpoint.
10. The method for coordinated optimization control of excitation and static frequency converters for pumped start-up of pumped storage units according to claim 8, characterized in that, In the online optimization process of step S2, the interior point method or the effective set method is used to solve the optimization problem, and the numerical computing library used is qpOASES or OSQP.
11. The method for coordinated optimization control of excitation and static frequency converters for pumped start-up of pumped storage units according to claim 8, characterized in that, In step S2, the model prediction process adopts the linear time-varying MPC method, which linearizes the joint prediction model around the current operating point in each sampling period to obtain a linear state-space model (A,B,C,D).
12. The method for coordinated optimization control of excitation and static frequency converters for pumped start-up of pumped storage units according to claim 8, characterized in that, In step S3, during the repetition of step S2, a state estimation step is also included, in which a Kalman filter is used to estimate the state to eliminate noise and the load torque T_load is observed in real time.
13. The method for coordinated optimization control of excitation and static frequency converters for pumped start-up of pumped storage units according to claim 8, characterized in that, In step S4, when the unit speed reaches 95%-100% of the synchronous speed and the power angle is stable, the system issues a "synchronous grid connection" command, SFC exits, and the excitation system switches to normal voltage regulation mode.