Method and system for voltage compensation of double-parallel bus based on model predictive control

Abstract

The application relates to a model prediction control-based double-machine parallel bus voltage compensation method and system, and belongs to the technical field of voltage stability control of power systems, wherein the method comprises the following steps: adjusting the output power of two gas turbine generators according to the respective droop coefficients of the two gas turbine generators which are parallel to a direct-current bus; obtaining the current voltage deviation of the direct-current bus and the current output current values of the two gas turbine generators, and constructing current state variables; combining a state space model, predicting a voltage deviation sequence of the direct-current bus at multiple future time steps according to the current state variables and a preset reference voltage compensation sequence of the two gas turbine generators; optimizing the preset reference voltage compensation sequence according to the voltage deviation sequence, and adjusting the output voltage of the two gas turbine generators again according to a target reference voltage compensation sequence obtained through optimization. The application can significantly reduce the steady-state deviation of the bus voltage while basically maintaining the original steady-state power distribution ratio.

Description

Technical Field

[0001] This invention relates to the field of power system voltage stability control technology, and in particular to a method and system for voltage compensation of a dual-machine parallel bus based on model predictive control. Background Technology

[0002] In a dual-generator parallel power supply system, droop control is typically employed to achieve reasonable power distribution. In a DC generator system, droop control is essentially achieved by adjusting the generator excitation current. The principle is as follows: by detecting the deviation of the DC bus voltage, the reference voltage of the excitation system is adjusted, thereby adjusting the excitation current output by the excitation system. In response to the excitation current, the generator adjusts its output electromotive force and output power. In its specific implementation, by setting different droop coefficients (D) for the two generators, the excitation systems of each generator can generate different amplitudes of excitation current adjustment to the same bus voltage deviation, thus controlling the power distribution ratio between the two generators.

[0003] However, droop control schemes for related DC systems generally suffer from inherent defects in control coupling: to ensure power distribution accuracy, voltage steady-state error suppression is sacrificed (frequency modulation leads to a decrease in bus voltage accuracy); or, to compensate for voltage deviation, the droop coefficient is modified, compromising power distribution accuracy. Related technologies have yet to find a decoupled control scheme that can simultaneously maintain the basic power distribution characteristics of droop control and significantly reduce the steady-state deviation of the bus voltage. Summary of the Invention

[0004] In view of this, it is necessary to provide a method and system for compensating bus voltage in a dual-machine parallel system based on model predictive control, in order to solve the problem that related technologies have never had a decoupled control scheme that can simultaneously maintain the basic characteristics of droop control power distribution and significantly reduce the steady-state deviation of bus voltage.

[0005] To address the aforementioned problems, in a first aspect, the present invention provides a method for voltage compensation of a dual-machine parallel bus based on model predictive control, comprising:

[0006] The output power of the two gas turbine generators is adjusted according to their respective droop coefficients connected in parallel to the DC bus; wherein the ratio of the droop coefficients of the two gas turbine generators is inversely proportional to the rated power of the two gas turbine generators.

[0007] The current voltage deviation of the DC bus and the current output current values ​​of the two gas turbine generators are obtained to construct the current state variable; based on the current state variable and the preset reference voltage compensation sequence of the two gas turbine generators, combined with the state space model describing the dynamic relationship between the reference voltage compensation and the DC bus voltage deviation, the voltage deviation sequence of the DC bus in the future multiple time steps is predicted.

[0008] The preset reference voltage compensation sequence is optimized based on the voltage deviation sequence to obtain the target reference voltage compensation sequence, and the output voltages of the two gas turbine generators are adjusted again based on the target reference voltage compensation sequence.

[0009] Secondly, the present invention also provides a model predictive control-based dual-machine parallel bus voltage compensation system, comprising:

[0010] DC bus;

[0011] Two gas turbine generators connected in parallel on the DC bus;

[0012] The control unit is communicatively connected to the two gas turbine generators and is used to implement the model predictive control-based dual-machine parallel bus voltage compensation method described above.

[0013] Thirdly, the present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the model predictive control-based dual-machine parallel bus voltage compensation method described in any of the preceding claims.

[0014] The beneficial effects of this invention are:

[0015] This invention does not modify the droop coefficient, but adjusts the output power of the two gas turbine generators based on a fixed droop coefficient, ensuring power distribution accuracy. Simultaneously, considering that frequency regulation will lead to bus voltage deviation, the current voltage deviation of the DC bus and the current output current values ​​of the two gas turbine generators are further obtained to construct the current state variable. Based on the current state variable and the preset reference voltage compensation sequence of the two gas turbine generators, combined with the state-space model describing the dynamic relationship between reference voltage compensation and DC bus voltage deviation, the voltage deviation sequence of the DC bus for multiple future time steps is recursively predicted. The preset reference voltage compensation sequence is optimized based on the voltage deviation sequence to obtain the target reference voltage compensation sequence, and the output voltage of the two gas turbine generators is adjusted again based on the target reference voltage compensation sequence, thereby reducing the bus voltage deviation. Furthermore, according to theoretical derivation and experimental verification, the bus voltage correction based on voltage compensation has virtually no impact on the steady-state power distribution of the previous step. This invention achieves control decoupling, simultaneously maintaining the droop control power distribution characteristics and significantly reducing the steady-state deviation of the bus voltage. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating an embodiment of the model predictive control-based dual-machine parallel bus voltage compensation method provided by the present invention.

[0018] Figure 2 The excitation current regulating droop control principle diagram provided by the present invention;

[0019] Figure 3 The optimized DC bus voltage control structure diagram for parallel operation of generator sets provided by this invention;

[0020] Figure 4 The logic diagram for MPC voltage deviation compensation provided by this invention;

[0021] Figure 5 The applied load waveform diagram provided by the present invention;

[0022] Figure 6 A comparison chart of bus voltages for 60-90kW propulsion loads provided by this invention;

[0023] Figure 7 A power comparison chart of two engines with a 60-90kW propulsion load provided for this invention;

[0024] Figure 8 Another applied load variation diagram provided by the present invention;

[0025] Figure 9 A power comparison chart of dual-machine propulsion loads of 30-60-90-60-30kW provided for this invention;

[0026] Figure 10 A comparison diagram of bus voltages for 30-60-90-60-30kW propulsion loads provided by the present invention;

[0027] Figure 11 This is a schematic diagram of an embodiment of the model predictive control-based dual-machine parallel bus voltage compensation system provided by the present invention. Detailed Implementation

[0028] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0030] In the description of the embodiments of this invention, unless otherwise stated, "a plurality of" means two or more. The terms "first," "second," etc., used in the embodiments of this invention are used to distinguish similar objects, and are not used to describe a specific order or sequence, nor to indicate or imply their relative importance or implicitly specify the number of indicated technical features. It should be understood that such data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class, and the number of objects is not limited; for example, a first object can be one or more.

[0031] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0032] Reference Figure 1 This diagram illustrates a flowchart of an embodiment of the model predictive control-based dual-machine parallel bus voltage compensation method provided by the present invention. The method includes:

[0033] S101, adjust the output power of the two gas turbine generators according to their respective droop coefficients connected in parallel to the DC bus; wherein, the ratio of the droop coefficients of the two gas turbine generators is inversely proportional to the rated power of the two gas turbine generators.

[0034] This embodiment can be applied to dual-machine parallel DC power supply systems, such as new energy ship power systems, offshore engineering microgrids, emergency backup power supplies, and port shore power systems.

[0035] Droop control correlates the DC bus voltage with the load current by setting a droop coefficient, and the correlation conforms to the following:

[0036]

[0037] in, This is the DC bus voltage. This is the reference value for the DC bus voltage. The droop coefficient is i, where i=1 and 2 are the unit numbers.

[0038] In droop control, the voltage deviation of the DC bus can be detected in real time. The voltage deviation refers to the difference between the actual measured value of the DC bus voltage and the set reference value. When a voltage deviation is detected, the output power of the two gas turbine generators is adjusted according to their respective droop coefficients.

[0039] In one example, the adjustment process is as follows: Figure 2 As shown, it includes: if the control unit detects that the actual measured value of the DC bus voltage is lower than the set reference value, it controls the excitation system to increase the excitation current according to the droop coefficient; the generator receives the excitation current and adjusts the output power according to the excitation current.

[0040] The output power of two gas turbine generators is inversely proportional to their droop coefficient and directly proportional to their rated power. That is, a gas turbine generator with a larger rated power will have a greater output power and can handle a larger load. For example, by setting... : =P2_rated:P1_rated=2 (Prated is the rated power of the unit, D is the droop coefficient of the unit, and 1 and 2 are the unit numbers), which forces the two units to distribute active power according to the rated power ratio, and the unit 2 bears twice the load of the unit 1.

[0041] S102, obtain the current voltage deviation of the DC bus and the current output current values ​​of the two gas turbine generators, and construct the current state variables; based on the current state variables and the preset reference voltage compensation sequence of the two gas turbine generators, and combined with the state space model describing the dynamic relationship between the reference voltage compensation and the DC bus voltage deviation, predict the DC bus voltage deviation sequence for multiple future time steps.

[0042] Specifically, the control unit acquires the actual voltage of the DC bus at the current time step. And based on the actual voltage and the preset reference voltage Calculate the voltage deviation at the current time step. .

[0043] Voltage deviation is defined as:

[0044]

[0045] The control unit also acquires the output current values ​​of the two gas turbine generators at the current time step. and And construct the current state variable:

[0046]

[0047] The control unit uses the current state variables As the initial state predicted by the state-space model, and combined with the preset reference voltage compensation sequence, the predicted state variables and the predicted voltage deviation of the DC bus are recursively calculated for multiple future time steps.

[0048] It should be noted that the current voltage deviation These are current state variables and serve as initial conditions for predicting future states; they are not control inputs to the state-space model. The control input to the state-space model is the reference voltage compensation control quantity. The model output is the predicted voltage deviation for future time steps. .

[0049] Therefore, in this embodiment, the accuracy of the DC bus voltage is further controlled through S102 and S103.

[0050] Specifically, in this step, the reference voltages corresponding to the two gas turbine generators are the same at the same time step.

[0051] The preset reference voltage compensation sequence includes reference voltage compensation at the current time step and at multiple future time steps.

[0052] S103, optimize the preset reference voltage compensation sequence according to the voltage deviation sequence to obtain the target reference voltage compensation sequence, and adjust the output voltage of the two gas turbine generators again according to the target reference voltage compensation sequence.

[0053] The preset reference voltage compensation sequence can be optimized to minimize voltage deviation, resulting in an optimized target reference voltage compensation sequence. This target reference voltage compensation is then superimposed on the excitation system's reference voltage. Based on the difference between the superimposed target reference voltage compensation and the actual terminal voltage of the gas turbine generator (the superimposed compensation value, equivalent to reverse compensation for the steady-state voltage drop caused by droop control), the excitation system's output excitation current is controlled, thereby achieving a secondary adjustment of the gas turbine generator's terminal voltage. This secondary adjustment reduces DC bus voltage deviation. (Refer to...) Figure 3 The diagram shows an optimized DC bus voltage control structure for parallel operation of generator sets provided by the present invention.

[0054] The model predictive control-based dual-machine parallel bus voltage compensation method provided in this embodiment can be applied to a model predictive control-based dual-machine parallel bus voltage compensation system, and more specifically, to the control unit in the system.

[0055] In summary, this embodiment does not modify the droop coefficient, but adjusts the output power of the two gas turbine generators based on a fixed droop coefficient, ensuring power distribution accuracy. Simultaneously, considering that frequency regulation will lead to bus voltage deviation, the current DC bus voltage deviation and the preset reference voltage compensation sequences of the two gas turbine generators are further used, combined with a state-space model describing the relationship between voltage deviation and voltage compensation, to predict the DC bus voltage deviation sequence for multiple future time steps. The preset reference voltage compensation sequence is then optimized based on the voltage deviation sequence to obtain the target reference voltage compensation sequence. The output voltage of the two gas turbine generators is then adjusted again based on the target reference voltage compensation sequence, thereby reducing the bus voltage deviation. Furthermore, both formula derivation and experimental verification demonstrate that the bus voltage correction based on voltage compensation has minimal impact on the previous steady-state power distribution.

[0056] Specifically, the derivation of the formula is as follows: Let the first... The sag relationship of the gas turbine generator is as follows:

[0057]

[0058] in, This is the DC bus voltage. The reference voltage, For the first Reference voltage compensation for the unit. The droop coefficient is... This is the output current.

[0059] If both generating units apply the same direction common compensation:

[0060]

[0061] Then we have:

[0062]

[0063] Therefore, the output current ratio of the two units is:

[0064]

[0065] Under steady-state conditions where the DC bus voltage is the same and the unit terminal voltage is approximately the same:

[0066]

[0067] so

[0068]

[0069] This means that as long as the secondary compensation uses the same amount of public compensation... The load is applied to both units in the same direction without changing the droop coefficient. , In this case, the steady-state power distribution ratio is still mainly determined by the droop coefficient ratio.

[0070] In summary, this embodiment achieves control decoupling, simultaneously maintaining the basic characteristics of droop control power distribution while significantly reducing the steady-state deviation of the bus voltage. Furthermore, it requires no modification to the existing hardware platform, excitation control main circuit, or existing droop control logic of the original dual-machine parallel system. Only a new voltage compensation module based on Model Predictive Control (MPC) needs to be added and integrated into the existing industrial controller. This module communicates with the gas turbine excitation system via CAN bus or Ethernet, thus completing the technology implementation.

[0071] In some embodiments of the present invention, S102 includes:

[0072] The dynamic relationship between the reference voltage compensation and the DC bus voltage deviation is described by the following discrete state-space model:

[0073]

[0074]

[0075] In the formula, This refers to the state variables at the current time step. This is the reference voltage compensation control value for the current time step; The state matrix; The input matrix; Select a matrix for voltage deviation.

[0076] The state variables are:

[0077]

[0078] in, This represents the DC bus voltage deviation at the current time step. and These are the output current values ​​of the two gas turbine generators at the current time step.

[0079] The voltage deviation selection matrix is ​​as follows:

[0080]

[0081] The voltage deviation selection matrix is ​​used to extract the DC bus voltage deviation from the state variables.

[0082] At the current time step With the current state variable As the initial state for prediction, and based on the preset reference voltage compensation sequence, the following formula is used to predict the first... State variables at each time step:

[0083]

[0084] in, , To predict the number of time steps that need to be predicted within the time domain; Indicates the current time step The set number Reference voltage compensation control quantity for the time step; Indicates the current time step The predicted first The state variables at each time step.

[0085] Based on the predicted state variables, the predicted DC bus voltage deviation for multiple future time steps is obtained using the following formula:

[0086]

[0087] in, Indicates the current time step The predicted first The DC bus voltage deviation at each time step. The prediction method can be: prediction based on simulation experiments using predicted state variables.

[0088] Actual voltage deviation at the current time step As part of the state variables, it is used to initialize state prediction; reference voltage compensation control quantity. For the control input of the state-space model; predict voltage deviation This is the predicted output of the state-space model for future time steps.

[0089] The reference voltage compensation control quantity affects the state variables of subsequent time steps through the state transition equation.

[0090] In a specific example, ,a is a coefficient reflecting the inertial simulation characteristics of the current, and its value can be 0.995; b is a coefficient characterizing the negative influence of the current on the voltage deviation, and its value can be -0.001 or -0.002. c is a coefficient characterizing the effect of the preset reference voltage compensation on the voltage deviation, and its value can be 1 or 1.8. One time step Ts can be 0.02s. It can be 8-20. For example, =8 indicates a predicted voltage deviation within the next 0.16 seconds.

[0091] In this embodiment, the input matrix is ​​used to manipulate only the state variables. It directly affects voltage deviation. The output matrix is ​​used to directly output state variables, and the output state variables are used for simulation feedback. Furthermore, the model in this embodiment does not include a direct feedthrough term where the control input at the same time step directly affects the output; that is... This formula indicates that the dual-machine compensation value does not directly affect the output voltage deviation at the same time step, but rather needs to go through the dynamic response of the gas turbine excitation, motor and bus system before affecting the bus voltage.

[0092] It has long been widely believed in the industry that MPC's rolling optimization involves large computational loads and poor real-time performance, making it only suitable for steady-state power scheduling and optimization in microgrids. It cannot adapt to dynamic voltage control scenarios requiring millisecond-level responses, such as load surges, and is even less suitable for rapid adjustment of gas turbine excitation systems. This embodiment, targeting the operating characteristics of a dual-machine parallel system, optimizes the prediction and control time domains (Np=20, Nc=3), constructing a low-order, accurate discrete state-space model. This significantly reduces the computational load, enabling rolling optimization to be solved within a 0.02s sampling period. Under load step surges, voltage deviation convergence can be achieved within two control cycles, making it suitable for dynamic compensation scenarios. This demonstrates the feasibility of MPC in DC power supply rapid control scenarios and overcomes this technical bias.

[0093] In some embodiments of the present invention, the step of optimizing and adjusting the reference voltage compensation sequence based on the voltage deviation sequence to obtain the target reference voltage compensation sequence includes:

[0094] Solve for the target reference voltage compensation sequence that minimizes the cost function; where the cost function is expressed as follows:

[0095]

[0096] In the formula, J represents the cost; k is the current time step number, i=1,2,…, ; The number of time steps that need to be predicted; ; The DC bus voltage deviation at the (k+i)th time step is predicted based on the current state variables and the preset reference voltage compensation sequence at the current time step k; u1(k+i-1) and u2(k+i-1) are the reference voltage compensations for the two gas turbine generators, respectively. u1(k+i-1) and u2(k+i-1) represents the reference voltage compensation rate of change for the two gas turbine generators, respectively. , as well as These are the weighting coefficients.

[0097] In a specific example, =[100,0.1,0.11] represents the voltage deviation penalty weight, which is used to prioritize stability accuracy while taking into account the smoothness of current prediction. , which represents the penalty weight for compensation, used to avoid overcompensation. =0.8 indicates the penalty weight for the rate of change of compensation amount, which is used to ensure smooth adjustment of the dual machines. It can be 3-5, for example, =4 indicates that the dual-machine compensation value is optimized within the next 0.08s.

[0098] In summary, this embodiment does not modify the droop coefficient in the primary droop control, nor does it change the proportional relationship between the droop coefficients of the two units. When the compensation is superimposed on the excitation reference voltages of the two units in the same direction, it is equivalent to shifting the overall droop characteristic curves of the two units without changing the droop slope. Therefore, the steady-state power distribution ratio remains essentially unchanged. Meanwhile, considering non-ideal factors such as excitation system dynamics, unit inertia, line impedance, and actuator limiting, secondary compensation may cause small power fluctuations during dynamic processes. This embodiment suppresses this effect through compensation amount constraints and compensation change rate constraints, and simulation results verify that the secondary compensation strategy has a minimal impact on the steady-state power distribution ratio.

[0099] In some embodiments of the present invention, the step of solving the target reference voltage compensation sequence when minimizing the cost function includes:

[0100] Based on preset constraints, the target reference voltage compensation sequence is solved when the cost function is minimized; wherein the preset constraints include: reference voltage compensation constraints, reference voltage compensation rate of change constraints, and voltage deviation constraints for multiple future prediction time steps.

[0101] In a specific example, the preset constraint is:

[0102]

[0103]

[0104]

[0105] The first constraint: physical limit of dual-machine compensation (actuator output range);

[0106] The second constraint: the rate of change of compensation (to avoid system oscillation caused by excessively rapid adjustment);

[0107] The third constraint: voltage deviation control target constraint.

[0108] In some embodiments of the present invention, the step of readjusting the output voltages of the two gas turbine generators again according to the target reference voltage compensation sequence includes:

[0109] The output voltages of the two gas turbine generators are adjusted again based on the first compensation value in the target reference voltage compensation sequence, and the steps are returned to predict the voltage deviation sequence of the DC bus for multiple future time steps.

[0110] Specifically, although the compensation values ​​corresponding to multiple time steps are optimized, only the compensation value of the first time step is used to adjust the output voltage of the two gas turbine generators, and then the process returns to S102 to continue prediction and optimization in the next time step.

[0111] In some embodiments of the present invention, the step of readjusting the output voltage of the two gas turbine generators again based on the first compensation value in the target reference voltage compensation sequence includes:

[0112] The first compensation value in the target reference voltage compensation sequence is superimposed on the reference voltage of the excitation system of the two gas turbine generators respectively;

[0113] Based on the two excitation system reference voltages with superimposed compensation values, the output voltages of the two gas turbine generators are adjusted again.

[0114] In some embodiments of the present invention, the model predictive control-based method for compensating the voltage of a dual-machine parallel bus further includes:

[0115] Obtain the actual voltage deviation of the DC bus in the next time step, and calculate the prediction error between the actual voltage deviation and the voltage deviation of the next time step predicted in the current time step;

[0116] Adjust the parameters in the model based on the prediction error.

[0117] For example, the values ​​of a, b, and c can be adjusted, or other parameters in the simulation model can also be adjusted.

[0118] Reference Figure 4 The diagram illustrates a logic diagram for implementing MPC voltage deviation compensation according to the present invention, including:

[0119] (1) Prediction stage: Based on the state-space model, the current voltage deviation is input. Dual-machine compensation amount Predicting the voltage deviation sequence for the next Np steps , ...,

[0120] (2) Optimization stage: Under constraints, solve for the dual-machine compensation sequence that minimizes the cost function J. This enables the coordinated allocation of compensation amounts between the two machines.

[0121] (3) Execution phase: Only the first step of the compensation in the optimization sequence is executed. Control the coordinated adjustment of the output of the two machines;

[0122] (4) Feedback correction stage: Collect the actual voltage of the DC bus and the actual output current of the two gas turbine generators in the next time step, and reconstruct the actual state variables in the next time step. ; Calculate the actual voltage deviation at the next time step The voltage deviation from the predicted voltage at the current time step The prediction error between the two is used to correct the model parameters and perform rolling prediction and optimization for the next control cycle.

[0123] To verify the effectiveness of this invention, a harsh operating condition was designed for comparison with the classic single-drop strategy. The experimental platform used a 60kW propulsion load as a baseline, and at t=9s, a sudden 30kW step disturbance was added, increasing the total load to 90kW. The load curve is shown below. Figure 5 As shown below, the comparison focuses on DC bus voltage stability and the dynamic characteristics of dual-machine output power.

[0124] 1. Comparison of bus voltages:

[0125] The bus voltage comparison under the two strategies is as follows: Figure 6 As shown:

[0126] It can be seen that the MPC compensation control restores the voltage to within 540±1V in steady state, which has a good control effect.

[0127] 2. Power comparison of the two machines:

[0128] To investigate the impact of MPC compensation control on the power of the dual-machine system, the power change curves of the dual-machine system before and after compensation were compared, such as... Figure 7 As shown in the figure, the MPC compensation strategy has little impact on power distribution under steady state.

[0129] Example 1: Simulation implementation under rated steady-state load conditions:

[0130] This embodiment verifies the basic scheme and corresponds to the simulation diagrams of steady-state voltage and steady-state power distribution in the instruction manual.

[0131] In a DC power supply simulation system consisting of a single-shaft gas turbine with a rated power of 100kW and a split-shaft gas turbine with a rated power of 200kW connected in parallel, the two gas turbines are connected to the 540V rated DC bus through their respective rectification and excitation systems, and the total load of the simulation system is set to a constant rated load of 60kW.

[0132] In the droop control layer, droop coefficients D1 and D2 are set for the two gas turbines respectively, so that D1:D2=200:100 (inversely proportional to the rated power of the two turbines), thereby achieving a 1:2 active power distribution under steady-state load, that is, the 100kW single-shaft gas turbine undertakes a 20kW load and the 200kW split-shaft gas turbine undertakes a 40kW load.

[0133] In the MPC compensation layer, based on the discrete state-space model (sampling time Ts=0.02s) identified by the simulation system, the bus voltage deviation is used as the only control object. The voltage deviation for the next 20 control steps is predicted. Under preset constraints, the dual-machine compensation sequence that minimizes the cost function is solved. Only the first compensation quantity of the sequence is superimposed in the same direction on the excitation reference voltage of the two gas turbines without changing the proportional relationship of the droop control.

[0134] Simulation results: Under steady-state conditions, the DC bus voltage of this embodiment remains stable within the range of 540V±1V; the output power of the two machines always maintains a 1:2 distribution ratio, and the power distribution accuracy is less affected by MPC compensation, thus verifying the effectiveness of the decoupling control of this invention.

[0135] Example 2: Simulation implementation under step load change condition:

[0136] This embodiment is a core dynamic performance verification, and corresponds to the load step waveform, bus voltage comparison, and dual-machine power comparison diagrams in the instruction manual.

[0137] Under the premise that the simulation system structure and control parameters remain unchanged in Example 1, the simulation load condition is set as follows: the initial steady-state load is 60kW, and at the simulation time t=9s, a step load disturbance is introduced within 50ms, and the total load jumps from 60kW to 90kW.

[0138] During load surges, the droop control layer first redistributes transient power by rapidly adjusting the excitation current, maintaining a 1:2 capacity allocation ratio between the two turbines. However, the bus voltage drops significantly as the load increases, with the maximum drop exceeding 10V. The MPC compensation layer identifies the voltage deviation trend in real time during the prediction phase and solves the dual-turbine collaborative compensation amount synchronously through rolling optimization, then adds the compensation amount to the excitation reference terminals of the two gas turbines.

[0139] Simulation results: After a sudden load change, this embodiment can bring the bus voltage deviation to within ±1V within 2 control cycles (0.04s), and the voltage recovery speed is more than 80% faster than the traditional simple droop control, reducing overshoot and oscillation. During the transient and steady-state processes, the two machines always maintain a 1:2 power distribution ratio, with no circulating current and no distribution misalignment problems, verifying the dynamic compensation effect of the present invention under the load change scenario.

[0140] Example 3: Simulation implementation under wide-range variable load cyclic conditions:

[0141] This embodiment is an extension verification, covering the adaptability of the full load range, further supporting the universality of the patented solution, and can directly correspond to the simulation diagrams of multiple load cycles.

[0142] Under the premise that the simulation system structure and control parameters remain unchanged in Example 1, the simulation variable load cycle condition is set: the total simulation time is 30s, 0-5s maintains 30kW light load, 5-10s jumps to 60kW rated load, 10-15s jumps to 90kW heavy load, 15-20s drops back to 60kW rated load, 20-25s drops back to 30kW light load, 25-30s no load, fully covering the full load range of no load-light load-rated load-heavy load.

[0143] Throughout the entire operating process, the droop control layer independently completes the power distribution between the two generators, synchronously adjusting the excitation current according to real-time load changes to maintain a 1:2 power distribution ratio. The MPC compensation layer constructs state variables based on the current voltage deviation and the current output current values ​​of the two gas turbine generators, and performs rolling optimization and compensation based only on the real-time bus voltage deviation. It does not interfere with the power distribution logic of the droop control, and achieves adaptive compensation for voltage deviation under different loads.

[0144] Simulation results: Under full load cyclic conditions, the DC bus voltage of this embodiment can recover to within 540V±2V within 3 control cycles during all load switching processes, and the maximum voltage deviation under all conditions does not exceed 2V; the power distribution error between the two machines is always less than 2% in the entire range from no load to heavy load, which verifies the stability and adaptability of the present invention under wide range of variable load scenarios.

[0145] This experiment simulates a multi-step load disturbance condition. The specific load power segments are: 30kW for 0–5s, 60kW for 5–10s, 90kW for 10–15s, 60kW for 15–20s, and 30kW for 20–25s. The load change points are 5s, 10s, 15s, and 20s. This experiment is used to test the dynamic response capability, power distribution rationality, and bus voltage stability of the MPC compensation control strategy and the individual droop control strategy during the rapid load increase and decrease process.

[0146] Experimental results are as follows Figure 8 , Figure 9 and Figure 10 As shown. Figure 8 The step change process of the propulsion load is presented. The load switching occurs at 5s, 10s, 15s and 20s respectively, and the power exhibits a segmented step characteristic. Figure 9The results of the power comparison between the two units show that during load changes, the two units using MPC compensation control can respond quickly and achieve reasonable power distribution. Compared with individual droop control, it has smaller overshoot and faster stabilization speed. Figure 10 The dynamic changes in bus voltage were demonstrated. During load switching, MPC compensation control effectively suppressed voltage fluctuations, allowing the bus voltage to quickly recover to a stable value, and significantly reducing voltage deviation. This verifies the superior performance of the proposed method in terms of power distribution coordination and voltage stability. The power comparison curves of the two machines show that both strategies can achieve matching between the total power of the two machines and the load power, meeting the system power balance requirements. Furthermore, MPC control has a relatively small impact on the power distribution between the two machines.

[0147] As can be seen from the bus voltage comparison curves, both strategies can maintain the bus voltage near the rated value, but MPC compensation control performs better in steady-state accuracy.

[0148] Steady-state voltage accuracy: MPC compensation control stabilizes the bus voltage at approximately 539V, which is closer to the rated 540V, with a voltage deviation of only 1V; the steady-state voltage of individual droop control is approximately 537V, with a deviation of approximately 3V, which deviates more from the rated value and has lower steady-state accuracy.

[0149] Reference Figure 11 This diagram illustrates a structural schematic of an embodiment of the model predictive control-based dual-machine parallel bus voltage compensation system provided by the present invention. The system includes:

[0150] DC bus;

[0151] Load module connected to the DC bus;

[0152] Two gas turbine generators connected in parallel on the DC bus;

[0153] The control unit is communicatively connected to the two gas turbine generators and is used to implement the model predictive control-based dual-machine parallel bus voltage compensation method mentioned above.

[0154] In this embodiment, the control unit can be connected to the excitation system of each of the two gas turbine generators and control the gas turbine generators through the excitation system.

[0155] In some embodiments of the present invention, the control unit includes:

[0156] The power distribution module is used to adjust the output power of the two gas turbine generators according to their respective droop coefficients connected in parallel to the DC bus; wherein the ratio of the droop coefficients of the two gas turbine generators is inversely proportional to the rated power of the two gas turbine generators.

[0157] The voltage compensation module is used to acquire the current voltage deviation of the DC bus and the current output current values ​​of the two gas turbine generators, constructing the current state variables. Based on the current state variables and the preset reference voltage compensation sequences of the two gas turbine generators, and combined with the state-space model describing the dynamic relationship between the reference voltage compensation and the DC bus voltage deviation, it predicts the DC bus voltage deviation sequence for multiple future time steps. The preset reference voltage compensation sequence is then optimized based on the voltage deviation sequence to obtain the target reference voltage compensation sequence, and the output voltages of the two gas turbine generators are adjusted again based on the target reference voltage compensation sequence. It should be noted that the implementation principles or processes of the above modules can refer to the aforementioned embodiment of the dual-generator parallel bus voltage compensation method based on model predictive control, and will not be elaborated upon here.

[0158] In one embodiment, the present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of any of the above-described model predictive control-based dual-machine parallel bus voltage compensation methods.

[0159] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware, and the program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.

[0160] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for compensating the voltage of a dual-machine parallel bus based on model predictive control, characterized in that, include: The output power of the two gas turbine generators is adjusted according to their respective droop coefficients connected in parallel to the DC bus; wherein the ratio of the droop coefficients of the two gas turbine generators is inversely proportional to the rated power of the two gas turbine generators. The current voltage deviation of the DC bus and the current output current values ​​of the two gas turbine generators are obtained to construct the current state variable; based on the current state variable and the preset reference voltage compensation sequence of the two gas turbine generators, combined with the state space model describing the dynamic relationship between the reference voltage compensation and the DC bus voltage deviation, the voltage deviation sequence of the DC bus in the future multiple time steps is predicted. The preset reference voltage compensation sequence is optimized based on the voltage deviation sequence to obtain the target reference voltage compensation sequence, and the output voltages of the two gas turbine generators are adjusted again based on the target reference voltage compensation sequence.

2. The method for compensating the voltage of a dual-machine parallel bus based on model predictive control according to claim 1, characterized in that, The step of predicting the DC bus voltage deviation sequence for multiple future time steps based on the current state variables and the preset reference voltage compensation sequences of the two gas turbine generators, combined with a state-space model describing the dynamic relationship between reference voltage compensation and DC bus voltage deviation, includes: The dynamic relationship between the reference voltage compensation and the DC bus voltage deviation is described by the following discrete state-space model: ； ； Based on the state variables at the current time step and the preset reference voltage compensation sequence, the predicted state variables for multiple future time steps are recursively obtained according to the following formula: ； The predicted voltage deviation of the DC bus at multiple future time steps is obtained according to the following formula: ； In the formula, This is the current time step number; , To predict the number of time steps that need to be predicted within the time domain; Let be the state variable at the current time step, and: ； in, This represents the DC bus voltage deviation at the current time step. and These are the output current values ​​of the two gas turbine generators at the current time step; This is the reference voltage compensation control value for the current time step; The state matrix; The input matrix; Choose a matrix for the voltage deviation, and: ； Indicates the current time step Based on the current state variables and the preset reference voltage compensation sequence, for the first... The predicted state variables are obtained by predicting the state variables at each time step. Indicates the current time step The predicted first DC bus voltage deviation at time step.

3. The method for compensating the voltage of a dual-machine parallel bus based on model predictive control according to claim 1, characterized in that, The step of optimizing the reference voltage compensation sequence based on the voltage deviation sequence to obtain the target reference voltage compensation sequence includes: Solve for the target reference voltage compensation sequence that minimizes the cost function; wherein the cost function is expressed as follows: ； In the formula, J represents the cost; k is the current time step number, i=1,2,…, ; The number of time steps that need to be predicted; ; The DC bus voltage deviation at the (k+i)th time step is predicted at the current time step k; u1(k+i-1) and u2(k+i-1) are the reference voltage compensations for the two gas turbine generators, respectively. u1(k+i-1) and u2(k+i-1) represents the reference voltage compensation rate of change for the two gas turbine generators, respectively. , as well as These are the weighting coefficients.

4. The method for compensating the voltage of a dual-machine parallel bus based on model predictive control according to claim 3, characterized in that, The target reference voltage compensation sequence for minimizing the cost function includes: Based on preset constraints, the target reference voltage compensation sequence is solved when the cost function is minimized; wherein, the preset constraints include: reference voltage compensation constraints, reference voltage compensation rate of change constraints, and voltage deviation constraints for multiple future prediction time steps.

5. The method for compensating the voltage of a dual-machine parallel bus based on model predictive control according to claim 1, characterized in that, The step of readjusting the output voltages of the two gas turbine generators again according to the target reference voltage compensation sequence includes: The output voltages of the two gas turbine generators are adjusted again based on the first compensation value in the target reference voltage compensation sequence, and the process returns to the step of predicting the voltage deviation sequence of the DC bus for multiple future time steps.

6. The method for compensating the voltage of a dual-machine parallel bus based on model predictive control according to claim 5, characterized in that, The step of readjusting the output voltages of the two gas turbine generators again based on the first compensation value in the target reference voltage compensation sequence includes: The first compensation value in the target reference voltage compensation sequence is superimposed on the reference voltage of the excitation system of each of the two gas turbine generators; Based on the two excitation system reference voltages with superimposed compensation values, the output voltages of the two gas turbine generators are adjusted again.

7. The method for compensating the voltage of a dual-machine parallel bus based on model predictive control according to claim 5, characterized in that, The method further includes: Obtain the actual voltage deviation of the DC bus in the next time step, and calculate the prediction error between the actual voltage deviation and the voltage deviation of the next time step predicted in the current time step; The parameters in the model are adjusted based on the prediction error.

8. A dual-machine parallel bus voltage compensation system based on model predictive control, characterized in that, include: DC bus; Two gas turbine generators connected in parallel on the DC bus; The control unit is communicatively connected to the two gas turbine generators and is used to implement the model predictive control-based dual-machine parallel bus voltage compensation method as described in any one of claims 1 to 7.

9. The system according to claim 8, characterized in that, The control unit includes: The power distribution module is used to adjust the output power of the two gas turbine generators according to the droop coefficients of the two gas turbine generators connected in parallel to the DC bus; wherein the ratio of the droop coefficients of the two gas turbine generators is inversely proportional to the rated power of the two gas turbine generators. The voltage compensation module is used to acquire the current voltage deviation of the DC bus and the current output current values ​​of the two gas turbine generators, and construct a current state variable; based on the current state variable and the preset reference voltage compensation sequence of the two gas turbine generators, and combined with the state space model describing the dynamic relationship between the reference voltage compensation and the DC bus voltage deviation, it predicts the voltage deviation sequence of the DC bus for multiple future time steps; it optimizes the preset reference voltage compensation sequence based on the voltage deviation sequence to obtain a target reference voltage compensation sequence, and adjusts the output voltage of the two gas turbine generators again based on the target reference voltage compensation sequence.

10. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the dual-machine parallel bus voltage compensation method based on model predictive control as described in any one of claims 1 to 7.

Images