A method and system for optimizing parameters of a power system stabilizer (PSS) of a thermal power unit in all operating conditions
By constructing a multi-condition PSS parameter optimization method, the problem of damping degradation of thermal power units under all operating conditions was solved, and the stability and adaptability of PSS parameters were improved over a wide load range, avoiding damping degradation and overcompensation, and enhancing the usability of the project.
Patent Information
- Application Number
- CN202611021743.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-09
- Publication Date
- 2026-11-10
AI Technical Summary
Existing PSS parameter tuning methods are difficult to adapt to scenarios such as low load, deep peak shaving, leading phase operation and weak power grid under all operating conditions of thermal power units, resulting in damping degradation and insufficient stability, and failing to provide sufficient positive damping.
A method for optimizing PSS parameters covering multiple operating conditions is constructed. The dominant electromechanical oscillation mode is obtained by establishing a linearization model and response identification. Dual-reference damping degradation constraints, phase compensation over-constraints, modal frequency offset constraints, and time-domain response constraints are introduced to optimize PSS parameters to meet the requirements of all operating conditions.
It improves the adaptability of PSS over a wide load operating range, avoids damping reduction and excessive phase compensation, enhances engineering usability and stability, and reduces the risk of polarity differences between the model and the actual device.
Smart Images

Figure CN123073775A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power system stability control technology, and in particular to a method and system for optimizing PSS parameters of thermal power units under all operating conditions. Background Technology
[0002] Power system stabilizers (PSS), as key additional damping control devices in synchronous generator excitation systems, are widely used to suppress low-frequency power oscillations in power systems and improve the stability of synchronous generators and power grids during small disturbances. In related technologies, PSSs collect signals such as speed deviation, frequency deviation, or power deviation, and through the coordinated operation of DC blocking, filtering, phase compensation, gain adjustment, and output limiting, inject additional stabilization control signals into the automatic voltage regulator to adjust the excitation voltage and generate additional damping torque. Specifically, traditional PSS parameter tuning methods are usually based on the uncompensated phase-frequency characteristics of the excitation system under rated or high-load conditions. These methods are implemented through on-site frequency characteristic tests, phase compensation tuning, gain tuning, and small-signal eigenvalue analysis, covering the entire process from parameter design to time-domain simulation verification.
[0003] However, with the increasing integration of high-proportion renewable energy sources and the growing demand for power system peak shaving, thermal power units frequently participate in deep peak shaving, rapid frequency regulation, and reactive power and voltage support, resulting in significant changes in their operating modes. Existing PSS parameter tuning methods directly adopt tuning parameters under high-load conditions, without fully considering changes in active and reactive power output, terminal voltage, and equivalent impedance of the external power grid under all operating conditions, including low load, deep peak shaving, leading phase operation, and weak grid conditions. This can lead to a mismatch between the amplitude-frequency and phase-frequency characteristics of the PSS compensation channel, failing to provide sufficient positive damping, and even causing localized damping degradation, thus affecting the grid-connected stability performance of thermal power units across the entire operating range. Summary of the Invention
[0004] The present invention aims to at least partially solve one of the technical problems in the related art.
[0005] The main objective of this invention is to provide a method for optimizing the PSS parameters of thermal power units under all operating conditions.
[0006] Another objective of this invention is to propose a full-condition optimization system for PSS parameters of thermal power units.
[0007] To achieve the above objectives, a first aspect of the present invention proposes a method for optimizing the PSS parameters of a thermal power unit under all operating conditions, comprising the following steps:
[0008] S1, construct a set of multiple operating conditions covering thermal power units under different active power output, reactive power output, terminal voltage, external power grid strength and operating mode; S2, for each operating condition in the set of multiple operating conditions, establish a linearized model of the synchronous generator, excitation system, power system stabilizer (PSS) and external power grid to obtain the dominant electromechanical oscillation mode, or obtain the dominant electromechanical oscillation mode through response identification, so as to establish a damping evaluation model under multiple operating conditions. S3. Construct a damping evaluation set that includes the state without a power system stabilizer, the original power system stabilizer parameter state, and the power system stabilizer parameter state to be optimized. Based on the damping evaluation set, construct a dual-benchmark damping degradation constraint to avoid damping degradation of the parameters to be optimized under certain operating conditions during the optimization process. S4. Based on the dual-benchmark damping degradation constraint, a comprehensive engineering constraint is constructed by combining at least one of the phase compensation over-constraint, modal frequency offset constraint, time-domain response availability constraint, parameter deviation constraint, and damping saturation constraint. A multi-condition comprehensive objective function is constructed and solved according to the comprehensive engineering constraint to obtain the power system stabilizer optimization parameters applicable to all operating conditions.
[0009] In one embodiment of the present invention, the process of constructing a set of multiple operating conditions includes: The system acquires operating data of thermal power units under at least one of the following operating conditions: rated or high load operation, low load operation, deep peak shaving operation, leading phase operation, operation under different external power grid strengths, and operation under different reactive power outputs. Based on this operating data, a set of multiple operating conditions is constructed, wherein the set of multiple operating conditions is represented as follows:
[0010] in, For the active power output of the unit under the i-th operating condition, To exert effort for no reason This refers to the terminal voltage or bus voltage. Equivalent impedance to the external power grid. This is the operating mode identifier, and N is the number of operating conditions.
[0011] In one embodiment of the present invention, the step of establishing a linearized model of the synchronous generator, excitation system, power system stabilizer, and external power grid for each operating condition in the set of multiple operating conditions to obtain the dominant electromechanical oscillation mode, or obtaining the dominant electromechanical oscillation mode through response identification, includes: For each operating condition, the dominant electromechanical oscillation mode parameters are obtained through at least one of the following methods: small-signal modal analysis, simulation response identification, field disturbance response identification, or online measurement data identification. The dominant electromechanical oscillation mode parameters include at least the damping ratio and the mode frequency.
[0012] In one embodiment of the present invention, the construction includes a damping evaluation set comprising the state without a power system stabilizer, the original power system stabilizer parameter state, and the power system stabilizer parameter state to be optimized, and a dual-benchmark damping degradation constraint is constructed based on the damping evaluation set, including: For each operating condition, calculate the dominant electromechanical mode damping ratio when the power system stabilizer is not engaged. Dominant electromechanical mode damping ratio when using original power system stabilizer parameters and the dominant electromechanical mode damping ratio when using the parameters of the power system stabilizer to be optimized. ; The key PSS parameter that significantly affects the damping of low-frequency oscillations was selected as the optimization variable, namely... The remaining parameters are fixed according to the original setting value of the excitation device or the typical engineering value; The dual-benchmark damping degradation constraint is constructed as follows: The formula is:
[0013]
[0014]
[0015] in, The dominant electromechanical mode damping ratio for the i-th operating condition without PSS. The dominant electromechanical mode damping ratio for the i-th operating condition under the original PSS parameter state. To determine the dominant electromechanical mode damping ratio for the i-th operating condition under the PSS parameter Θ to be optimized, and To allow for damping margin, The minimum damping threshold, Let be the weight of the i-th operating condition.
[0016] In one embodiment of the present invention, phase compensation over-constraint is used to limit the ratio of the time constant of the PSS compensation stage, the amplitude gain within the electromechanical oscillation frequency band, or the phase compensation deviation; the phase compensation over-constraint is expressed as follows:
[0017] in, For electromechanical oscillations, the frequency band of concern is... The allowable amplitude magnification factor, For the phase of the PSS compensation channel under the i-th operating condition, To compensate for the phase of the target, To allow for phase deviation, L is the number of the lead-lag series. To optimize the frequency response of the phase compensation stage under the PSS parameters, This represents the frequency response of the phase compensation circuit under the original PSS parameters. This refers to the amplitude gain constraint in the phase compensation over-constraint. This is a time constant ratio constraint in the phase compensation over-constraint. This refers to the phase deviation constraint in the phase compensation over-constraint.
[0018] In one embodiment of the present invention, a modal frequency offset constraint is used to limit the offset magnitude of the dominant electromechanical modal frequency relative to the original PSS parameter state under the PSS parameters to be optimized; the process of constructing the modal frequency offset constraint includes: Let the dominant electromechanical mode frequency of the i-th operating condition under the original PSS parameters be... The dominant electromechanical mode frequency under the PSS parameters to be optimized is Then the modal frequency offset constraint for:
[0019] in, The dominant electromechanical mode frequency under the original power system stabilizer parameters for the i-th operating condition. The dominant electromechanical mode frequency under the power system stabilizer parameters to be optimized; Parameter deviation constraints are used to limit the degree of deviation of the PSS parameters to be optimized from the original PSS parameters; Damping saturation constraint is used to reduce or stop further damping increases after the damping index reaches the preset engineering requirements. The damping saturation evaluation value is set as follows: ,in This is the upper limit of the damping reward.
[0020] In one embodiment of the present invention, a time-domain response availability constraint is used to evaluate the engineering availability of PSS parameters based on the active power first swing amplitude, oscillation convergence time, PSS output amplitude, PSS output rate of change, terminal voltage deviation, or number of oscillations; the process of constructing the time-domain response availability constraint includes: For the i-th operating condition, define the time-domain response index. :
[0021] A time-domain response penalty is constructed based on the time-domain response index and a preset acceptable threshold. , is represented as:
[0022] in, The active power initial swing amplitude, To shorten the convergence time, Output the maximum value for PSS. For PSS output limits, The number of oscillations. This represents the acceptable threshold for the time-domain response.
[0023] In one embodiment of the present invention, the multi-condition integrated objective function is expressed as:
[0024]
[0025]
[0026]
[0027] in, This is a penalty term for modal frequency shift. This is a phase deviation penalty term. As a time-domain response availability penalty, This is a penalty term for parameters deviating from their original values. , , to The weighting coefficient is used; the time-domain response availability penalty is constructed based on at least two of the following indicators: active power initial swing amplitude, oscillation convergence time, PSS output maximum amplitude, PSS output rate of change, maximum deviation of terminal voltage, and number of oscillations.
[0028] In one embodiment of the present invention, the method further includes: performing a self-verification of PSS injection polarity and parameter mapping before or after PSS parameter optimization, wherein the self-verification step includes: Define the polarity variable σ∈{PSS output injection polarity variable} 1, +1}, calculate the multi-condition damping level corresponding to the original PSS parameters under the two polarities respectively, and determine:
[0029] Where σ is the polarity variable injected into the PSS output; When the difference in minimum damping ratio between the two polarities exceeds a set threshold, σ As the polarity verification result of PSS model parameter mapping and engineering download; when the difference does not exceed the set threshold, a polarity uncertainty prompt is output, and further verification is performed through electromagnetic transient small disturbance test or field small disturbance test.
[0030] To achieve the above objectives, a second aspect of the present invention provides a full-condition optimization system for PSS parameters of thermal power units, comprising: The multi-condition generation module is used to construct a set of multiple operating conditions covering thermal power units under different active power output, reactive power output, terminal voltage, external power grid strength and operating modes. The damping evaluation module is used to establish a linearized model of the synchronous generator, excitation system, power system stabilizer (PSS), and external power grid for each operating condition in the set of multiple operating conditions to obtain the dominant electromechanical oscillation mode, or to obtain the dominant electromechanical oscillation mode through response identification, so as to establish a damping evaluation model under multiple operating conditions. The constraint construction module is used to construct a damping evaluation set that includes the state without a power system stabilizer, the original power system stabilizer parameter state, and the power system stabilizer parameter state to be optimized. Based on the damping evaluation set, a dual-benchmark damping degradation constraint is constructed to avoid damping degradation of the parameters to be optimized under certain operating conditions during the optimization process. The parameter optimization and output module is used to construct comprehensive engineering constraints based on the dual-benchmark damping degradation constraints, combined with at least one of phase compensation over-constraint, modal frequency offset constraints, time-domain response availability constraints, parameter deviation constraints, and damping saturation constraints. Based on the comprehensive engineering constraints, a multi-condition comprehensive objective function is constructed and solved to obtain power system stabilizer optimization parameters applicable to all operating conditions.
[0031] The embodiments of the present invention have the following beneficial effects: 1. This invention is designed for all operating conditions of thermal power units, rather than just for setting PSS parameters under rated or high load conditions, thereby improving the adaptability of PSS across a wide load operating range.
[0032] 2. This invention constructs a dual-benchmark damping degradation constraint for both the PSS-free state and the original PSS state, which can avoid the damping decrease caused by the optimization parameters under some deep-tuning conditions and solve the problem that traditional multi-condition averaging optimization may mask the degradation of local conditions.
[0033] 3. This invention introduces PSS lead-lag ratio constraints, amplitude gain constraints, and phase compensation window constraints, which can prevent the intelligent optimization algorithm from pushing the PSS parameters to the boundary, resulting in excessive phase compensation, high-frequency glitches, or abnormal amplification of the PSS output.
[0034] 4. This invention introduces a modal frequency offset constraint and a damping saturation reward mechanism, which can avoid abnormally high oscillation frequency or excessive parameters due to the pursuit of excessively high small-signal damping, thereby improving the conservatism and engineering interpretability of the optimization parameters.
[0035] 5. This invention introduces small-signal-time domain joint constraints, which can unify the Heffron-Phillips small-signal model, electromagnetic transient simulation and field disturbance test into the parameter screening process, thereby improving the engineering usability of the optimized parameters.
[0036] 6. The present invention sets up a PSS injection polarity and parameter mapping self-verification mechanism, which can reduce the risk of negative damping caused by differences in polarity, per-unit reference and module implementation between the model, simulation platform and actual excitation device.
[0037] 7. This invention is not only applicable to the PSS2B model, but can also be extended to PSS2A, PSS4B and other power system stabilizers containing phase compensation links; it is not only applicable to single-machine infinite bus systems, but can also be extended to multi-machine systems, weak grid transmission systems and high renewable energy operation scenarios. Attached Figure Description
[0038] The above-described and additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, in which: Figure 1 A flowchart illustrating a full-condition optimization method for PSS parameters of a thermal power unit, provided as an embodiment of the present invention; Figure 2 This is a structural diagram of a full-condition optimization system for PSS parameters of a thermal power unit provided in an embodiment of the present invention. Detailed Implementation
[0039] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. Furthermore, the small-signal model, response identification method, optimization algorithm, and operating condition division method in this embodiment are merely examples and do not limit the scope of protection of the present invention. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.
[0040] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0041] A power system stabilizer (PSS) is an additional damping control device in the excitation system of a synchronous generator. It typically collects signals such as generator speed deviation, frequency deviation, power deviation, or combinations thereof, and after passing through DC blocking, filtering, phase compensation, gain adjustment, and output limiting, inputs the additional stabilization control signal to the automatic voltage regulator. This regulates the excitation voltage and generates additional damping torque to suppress low-frequency power oscillations in the power system and improve the small disturbance stability of the synchronous generator and the power system.
[0042] In traditional power systems, thermal power units mostly operate at or near rated load conditions, and the power oscillation mode (PSS) parameters are typically tuned based on the uncompensated phase-frequency characteristics of the excitation system under high active power output conditions. Existing PSS tuning methods generally include field frequency characteristic tests, phase compensation tuning, gain tuning, small-signal eigenvalue analysis, time-domain simulation verification, and parameter optimization methods based on intelligent optimization algorithms such as particle swarm optimization and genetic algorithms. These methods typically aim to improve the damping ratio of the dominant electromechanical oscillation mode under the target operating condition, meet phase compensation requirements, or improve the power oscillation attenuation effect after disturbances.
[0043] With the increasing integration of high-proportion renewable energy sources and the growing demand for peak shaving in the power system, the operating modes of thermal power units have undergone significant changes. Thermal power units no longer operate under a single high-load condition for extended periods, but instead frequently participate in deep peak shaving, rapid frequency regulation, and reactive power and voltage support. Some thermal power units have already demonstrated the ability to operate stably for extended periods under low-load or even deep peak shaving conditions. Under these conditions, the unit's active power output, reactive power output, terminal voltage, excitation system operating point, equivalent impedance of the external power grid, and the amplitude and phase relationship of the PSS input signal may all change significantly, thereby altering the amplitude-frequency characteristics, phase-frequency characteristics, and damping contribution of the PSS compensation channel.
[0044] For accelerated power type PSS or PSS2B type stabilizers, their damping effect depends not only on the PSS gain and the lead-lag phase compensation time constant, but also on the power channel filtering, the calculation method of the speed or frequency input signal, the notch filter, the output limiting, and the interaction characteristics with the excitation system and the external power grid. Under the conditions of wide load operation, leading phase operation, weak grid operation, and deep peak shaving operation of thermal power units, the PSS parameters originally tuned based on high load conditions may be difficult to continue to accurately match the uncompensated phase frequency characteristics of the current system, thus affecting the PSS's ability to provide positive damping to the electromechanical oscillation circuit.
[0045] Currently, engineering methods typically employ small-signal models, state-space eigenvalue analysis, electromagnetic transient simulation, or on-site disturbance response identification to evaluate the impact of PSS parameters on system damping. For example, a Heffron-Phillips small-signal model can be established based on the synchronous generator, excitation system, PSS, and external power grid to determine the damping ratio and modal frequency of the dominant electromechanical modes at different operating points. Alternatively, electromagnetic transient simulation platforms such as PSCAD or on-site disturbance recordings can be used to verify indicators such as the initial amplitude of active power swing, number of oscillations, convergence time, and PSS output amplitude. These methods provide an important foundation for PSS parameter tuning, but further improvements are still possible when considering the full-condition operation of thermal power units, especially deep peak-shaving conditions.
[0046] Specifically, the existing technology has the following drawbacks: Existing PSS parameter tuning and optimization methods can achieve good damping improvement under normal high-load operation scenarios, but they still have the following shortcomings under all operating conditions such as deep peak shaving of thermal power units, low load, leading phase operation and weak grid.
[0047] First, existing methods mostly focus on rated load or high active power output conditions as the primary setting targets, lacking a unified evaluation of various operating modes such as low load, deep peak shaving, leading phase operation, and different external grid strengths. Since changes in the operating point of thermal power units alter the uncompensated phase frequency characteristics of the excitation system, the amplitude of the PSS input signal, and the equivalent impedance of the external network, PSS parameters set according to high load conditions may not continue to provide sufficient positive damping under deep adjustment conditions. In fact, they may even cause the additional electromagnetic torque of the PSS to deviate from the positive damping direction, resulting in a weakening of the damping enhancement effect or damping degradation.
[0048] Second, existing multi-condition PSS parameter optimization methods typically focus on maximizing the damping ratio of a single condition, the average damping ratio across multiple conditions, or the minimum damping ratio. They lack explicit constraints on ensuring that "the system after PSS implementation is not worse than the system without PSS" and that "optimized parameters are not worse than the original PSS parameter state." In some deeply tuned conditions, the system itself may already possess a certain level of damping. If only the average damping index is pursued, the optimization algorithm may sacrifice the original stability of the local deeply tuned conditions in exchange for improving the overall evaluation index, thereby risking a decrease in damping in local conditions.
[0049] Third, existing intelligent optimization methods tend to push the PSS lead-lag compensation time constant, gain, or other parameters to their limits. Although such parameter combinations may appear as an increase in damping ratio in small-signal eigenvalue analysis, they may cause problems such as excessive phase compensation, abnormal amplification of PSS output amplitude, output glitches, increased initial swing amplitude, increased oscillation frequency, or abnormal upward shift of the dominant electromechanical mode frequency in electromagnetic transient simulation or actual excitation devices, resulting in insufficient engineering usability of the optimized parameters.
[0050] Fourth, existing PSS parameter optimization methods typically assume that the parameter mapping relationship between the simulation model, control model, and field excitation device is consistent, lacking a self-checking mechanism for differences in PSS output injection polarity, power channel sign, speed signal calculation method, and internal module implementation. When there are inconsistencies in polarity or per-unit references between the small-signal model, electromagnetic transient model, or actual excitation device, the same set of PSS parameters may change from positive damping behavior to negative damping behavior, increasing the risks in engineering installation and field application.
[0051] Therefore, existing technologies lack a PSS parameter optimization method for all operating conditions of thermal power units that can explicitly avoid damping degradation in deep-conditioning, suppress overcompensation and abnormal mode frequency shifts while covering high load, low load, deep peak shaving, leading phase operation and different grid intensity conditions, and perform engineering verification of the parameter mapping relationship between the small-signal model, electromagnetic transient model and actual excitation device.
[0052] This invention aims to address the problem that existing PSS parameter tuning methods for thermal power units are difficult to adapt to all operating conditions, especially the potential for damping degradation and insufficient engineering adaptability of traditional high-load operating condition tuning parameters under low-load, deep peak shaving, leading phase operation, and weak grid scenarios. Traditional PSS parameters are mostly tuned based on rated load or high active power output conditions. When thermal power units enter low-load, deep peak shaving, leading phase operation, or weak grid operation scenarios, the generator operating point, the uncompensated phase frequency characteristics of the excitation system, the amplitude of the PSS input signal, and the equivalent impedance of the external network all change. The original PSS parameters may no longer provide effective positive damping, and may even cause damping degradation under some deep-shaving conditions. The technical objective of this invention is to construct a PSS parameter optimization method and system that primarily aims to improve damping under multiple operating conditions, explicitly constrains deep-shaving conditions to prevent degradation, avoids excessive phase compensation, prevents excessive upward shift of modal frequencies, maintains acceptable time-domain response, and ensures consistent mapping between model and device parameters.
[0053] To address the aforementioned problems, this invention proposes a method and system for optimizing the PSS parameters of thermal power units under all operating conditions. This method aims to improve damping adaptability under all operating conditions, and establishes a PSS parameter optimization model covering multiple operating modes, constrained by the principles of "no degradation under deep-adjustment conditions, usability over a wide load range, and achievable engineering implementation." This invention obtains key PSS parameters applicable to the entire operating range of thermal power units by constructing dual-benchmark damping degradation constraints, phase compensation over-constraints, modal frequency shift constraints, time-domain response constraints, and a PSS injection polarity self-verification mechanism.
[0054] The following describes, with reference to the accompanying drawings, a method and system for optimizing the PSS parameters of a thermal power unit under all operating conditions, according to an embodiment of the present invention.
[0055] Example 1 This embodiment provides a method for optimizing the PSS parameters of thermal power units under all operating conditions. For example... Figure 1 As shown, the method includes the following steps: S1 constructs a set of multiple operating conditions covering thermal power units under different active power output, reactive power output, terminal voltage, external power grid strength, and operating modes.
[0056] This embodiment first constructs a set of multiple operating conditions covering different operating states of thermal power units to comprehensively reflect the dynamic operating characteristics of thermal power units under different load levels, different power grid environments, and different operating modes, providing a basis for subsequent analysis and optimization. Specifically, historical operating data, dispatching operating data, and simulation operating data of thermal power units during actual operation are acquired. Key operating parameters under different operating states are statistically analyzed and organized. Based on factors such as active power output, reactive power output, terminal voltage, external power grid strength, and operating mode, a set of multiple operating conditions covering various typical operating states is constructed. Among them, historical operating data can come from the unit's distributed control system, plant monitoring system, synchronous phasor measurement device, dispatch automation system, and production management system; simulation operating data can be generated through power system electromechanical transient simulation platform, electromagnetic transient simulation platform, or digital twin system to supplement extreme operating conditions that are difficult to obtain in field operation.
[0057] The multi-operating condition set constructed in this embodiment is represented as follows:
[0058] in, The active power output of the unit under the i-th operating condition; To exert effort for no gain; This refers to the terminal voltage or bus voltage. It is the equivalent impedance of the external power grid, used to characterize the strength of the unit's connection to the power grid; This is the operating mode identifier, used to describe the current operating mode of the unit; N is the number of operating conditions.
[0059] Among them, those who have contributed their efforts The preferred operating conditions cover the area near the rated capacity of thermal power units, the high-load operating range, the medium-load operating range, the low-load operating range, and the deep peak-shaving operating range, to ensure complete coverage of the system's dynamic characteristics under different load levels. For example, corresponding operating conditions can be established according to multiple output levels such as 100%, 90%, 80%, 70%, 60%, 50%, 40%, 30%, and 20% of the rated capacity of the units. Alternatively, a more refined operating condition distribution can be obtained by using continuous sampling based on actual dispatching and operation conditions.
[0060] No work done The configuration is based on the excitation system regulation status and reactive power regulation capability of the unit, including both normal lagging phase operation and leading phase operation, thereby covering the operating status of the unit under different reactive power support capabilities.
[0061] Terminal voltage The settings are based on the actual operating voltage level and can cover multiple voltage levels near the rated voltage and within the allowable fluctuation range to reflect the impact of different voltage operating conditions on the unit's wideband oscillation characteristics.
[0062] External power grid equivalent impedance This describes the electrical strength of the power grid to which the generating unit is connected. It can be calculated based on the power grid topology, short-circuit capacity, or equivalent network parameters. When the equivalent impedance of the external power grid is low, it indicates that the system's short-circuit capacity is high and the power grid is strong. When the equivalent impedance of the external power grid increases, it indicates that the system's short-circuit capacity is low, and the power grid is gradually transitioning to a weaker operating state. In actual implementation, operating conditions under different power grid strengths can be constructed by changing the equivalent impedance of the external network, the short-circuit ratio, or the parameters of the network equivalent model.
[0063] Operating mode identifier Used to identify different types of operating states, these include not only normal operating modes of the unit, but also typical operating modes resulting from changes in different operating environments and system structures. For example, It can correspond to various operating conditions, including rated load operation, high load operation, low load operation, deep peak shaving operation, leading phase operation, line disconnection operation, maintenance operation, weak grid operation caused by high output from new energy sources, peak shaving and frequency regulation superimposed operation, operation with different combinations of multiple transmission lines being put into operation and taken out of operation, and operation under different switching modes of reactive power compensation devices. For regional power grids with a high proportion of new energy sources, it can further consider typical scenarios such as large-scale grid connection of new energy sources, reduced system inertia, and decreased grid short-circuit capacity to enhance the coverage of the multi-operation condition set for the operating characteristics of new power systems.
[0064] Furthermore, to improve the representativeness of subsequent analysis results, each operating condition in the multi-operating condition set can be generated statistically from actual operating data, or automatically generated based on preset parameter combinations. When using the parameter combination method, active power output, reactive power output, generator terminal voltage, equivalent impedance of the external power grid, and operating mode can be discretized and values can be obtained separately. By permutation and combination, a multi-operating condition set covering the entire operating space can be generated. For cases with a large number of operating conditions, stratified sampling, Latin hypercube sampling, or random sampling methods can be used to select representative operating conditions, ensuring coverage of the operating space while reducing the complexity of subsequent calculations.
[0065] By constructing the above-mentioned set of multiple operating conditions, the typical operating states of thermal power units under different load levels, different grid intensities and different operating modes can be fully covered, providing a unified and complete data foundation for subsequent broadband oscillation characteristic analysis, key parameter identification, risk assessment and control strategy formulation.
[0066] S2, for each operating condition in the set of multiple operating conditions, establish a linearized model of the synchronous generator, excitation system, power system stabilizer (PSS), and external power grid to obtain the dominant electromechanical oscillation mode, or obtain the dominant electromechanical oscillation mode through response identification, so as to establish a damping evaluation model under multiple operating conditions.
[0067] In this embodiment, for the set of multiple operating conditions constructed in step S1, a corresponding small signal analysis model is established for each operating condition to obtain the dominant electromechanical oscillation mode of the thermal power unit under each operating condition. Based on the oscillation damping characteristics corresponding to each operating condition, a multi-operating condition damping evaluation model is established to provide an evaluation basis for subsequent broadband oscillation risk identification and control parameter optimization.
[0068] Specifically, for any operating condition in the set of multiple operating conditions, the active power output, reactive power output, generator terminal voltage, equivalent impedance of the external power grid, and operating mode corresponding to the operating condition are used as model input parameters to establish a linearized small-signal model composed of the synchronous generator, excitation system, power system stabilizer, and external power grid under the operating condition.
[0069] Preferably, the synchronous generator can be modeled using a classical second-order, third-order, fourth-order, or sixth-order model. Alternatively, a higher-order synchronous generator model incorporating damping windings, excitation windings, and flux linkage dynamics can be used as needed. The excitation system can employ IEEE standard excitation models, such as IEEE Type-ST1A, IEEE Type-AC1A, and IEEE Type-DC1A. The PSS can use IEEE standard PSS models, such as PSS1A, PSS2A, and PSS4B, with input signals selectable as speed deviation, frequency deviation, active power, or power angle variation. The external power grid is represented using the Thevenin equivalent model, with its equivalent parameters adjusted according to the system topology, short-circuit capacity, and network operation mode under the current operating conditions to reflect the system's dynamic characteristics under different grid intensities. It is important to emphasize that the aforementioned models and identification methods are only used to obtain damping evaluation parameters under different operating conditions. The focus of this invention is not on the specific model form, but rather on multi-condition three-state damping evaluation, dual-benchmark non-deterioration constraints, and comprehensive engineering verification mechanisms.
[0070] In a typical implementation, a Heffron-Phillips small-signal model is established based on the dynamic relationship between the synchronous generator, excitation system, PSS, and external power grid, and each component is uniformly represented in state-space form. Subsequently, the eigenvalues of the state matrix are solved to obtain all system eigenvalues and corresponding eigenvectors. At the same time, the contribution of each state variable to different oscillation modes can be analyzed based on the participation factor analysis. The dominant electromechanical oscillation mode with the greatest impact on system stability is identified from all oscillation modes, and its corresponding oscillation frequency, damping ratio, participation factor, and other characteristic parameters are extracted.
[0071] In another implementation, the dominant electromechanical oscillation mode is identified directly using system response data, without relying on a small-signal mathematical model. When electromagnetic transient simulation waveforms, real-time monitoring data, or disturbance recording data have been acquired on-site, the electromechanical oscillation characteristics can be directly identified based on the system response signal. The response signal may include dynamic signals such as generator speed, power angle, active power, reactive power, bus voltage, line power flow, electromagnetic torque, and frequency variation.
[0072] Preferably, the Prony analysis method can be used to fit the oscillation signal after the disturbance with exponential decay, and the system response can be expressed as the superposition of multiple damped oscillation components. The frequency, damping coefficient and initial amplitude of each oscillation component can be estimated, thereby obtaining the damping ratio of each dominant oscillation mode.
[0073] Furthermore, the matrix pencil method can be used to identify parameters of the acquired discrete response signals. By constructing a Hankel matrix and performing singular value decomposition, high-precision estimation of multiple damped oscillation mode parameters can be achieved. Alternatively, the least squares fitting method can be used to fit the damped oscillation model to the disturbance response curve, calculating the corresponding damping coefficient and oscillation frequency. The system damping ratio can also be calculated using the damping oscillation curve envelope identification method based on the attenuation law of the oscillation envelope curve. These identification methods can be used individually or in combination, and cross-validation through multiple algorithms can improve the accuracy and reliability of the damping parameter identification results.
[0074] For each operating condition in the set of multiple operating conditions, the above modeling analysis or response identification process is repeated to obtain the dominant electromechanical oscillation mode parameters under the corresponding operating condition, thereby establishing a damping evaluation model under multiple operating conditions.
[0075] S3. Construct a damping evaluation set that includes the state without a power system stabilizer, the original power system stabilizer parameter state, and the power system stabilizer parameter state to be optimized. Based on the damping evaluation set, construct a dual-benchmark damping degradation constraint to avoid damping degradation of the parameters to be optimized under certain operating conditions during the optimization process.
[0076] After obtaining the dominant electromechanical oscillation modes and damping evaluation results under multiple operating conditions, this embodiment further constructs a damping evaluation set that includes the state without a power system stabilizer, the original power system stabilizer parameter state, and the power system stabilizer parameter state to be optimized. Based on this damping evaluation set, a dual-benchmark damping degradation constraint is constructed to constrain the direction of damping improvement of the PSS parameter to be optimized under different operating conditions, so as to avoid the optimization process only improving the overall average damping level and causing the damping performance to deteriorate in some key operating conditions, especially the low-load deep peak shaving operating conditions.
[0077] Specifically, for the i-th operating condition in the set of multiple operating conditions Q Three PSS operating states were set up. The first state is the state without PSS, that is, in this operating condition, the power system stabilizer is not activated, only the dynamic models of synchronous generators, excitation systems and external power grids are retained, and the damping ratio of the dominant electromechanical oscillation mode is calculated in this state. The second method is the original PSS parameter state, which uses the PSS setting parameters already in operation in the existing excitation device of the power plant to calculate the damping ratio of the dominant electromechanical oscillation mode under this state. The third type is the state of PSS parameters to be optimized, which is the PSS parameter vector to be optimized. Substitute into the PSS model to calculate the damping ratio of the dominant electromechanical oscillation mode under this state. .
[0078] Among them, the state without PSS is used to characterize the basic damping level of the unit without additional damping control; the original PSS parameter state is used to characterize the actual damping level under the current engineering tuning parameters; and the PSS parameter state to be optimized is used to characterize the damping performance corresponding to the candidate parameters during the optimization process. By introducing the above three states simultaneously, the improvement effect of PSS parameters relative to the original engineering parameters can be considered during the optimization process, while avoiding the optimized parameters being lower than the state without PSS or lower than the minimum damping requirement of the system under local operating conditions.
[0079] In this embodiment, the PSS parameter vector This can include PSS gain, lead-lag phase compensation time constant, power channel filter time constant, notch filter parameters, and output limiting parameters. In actual implementation, the composition of optimization variables can be determined based on the specific structure of the excitation system and the PSS. When the PSS model includes multiple stages of lead-lag, DC blocking, filtering, notch filtering, and output limiting, parameters from these stages that have a significant impact on electromechanical oscillation damping can be included in the optimization variables.
[0080] In the preferred embodiment, to reduce the optimization dimensionality, improve engineering feasibility, and avoid impacting the original excitation device protection and limiting logic, only the key phase compensation parameters of the PSS that significantly affect the damping of low-frequency electromechanical oscillations are selected as optimization variables. For example, for a PSS containing two stages of lead-lag phase compensation, the vector of parameters to be optimized can be set as follows: ,in These represent the time constants in the two-stage lead-lag phase compensation circuit. Other parameters, such as PSS gain, DC blocking time constant, power measurement filter time constant, notch filter parameters, output upper and lower limits, and input signal filtering parameters, can be kept constant based on the excitation device's original settings, manufacturer-recommended values, or typical engineering values. This approach reduces the number of optimization variables while ensuring PSS damping adjustment capability, improving the interpretability and feasibility of parameter optimization results in practical engineering.
[0081] Furthermore, for each operating condition, calculations are performed based on the small-signal model or response identification method in step S2. , and A damping evaluation set is constructed. This damping evaluation set can be represented as a data set consisting of different operating conditions and their corresponding three types of damping results, used to characterize the damping differences between no PSS, original PSS parameters, and PSS parameters to be optimized under the same operating condition.
[0082] To avoid damping degradation under certain operating conditions after PSS parameter optimization, this embodiment does not use the original PSS parameter state as the sole comparison benchmark. Instead, it introduces both the no-PSS state and the original PSS parameter state as dual benchmarks, and combines them with the lowest damping threshold to form the condition-specific damping benchmark value. Specifically, for the i-th operating condition, the damping benchmark value is constructed as follows: in, The dominant electromechanical mode damping ratio for the i-th operating condition without PSS. The dominant electromechanical mode damping ratio for the i-th operating condition under the original PSS parameter state. and These represent the damping margin allowed relative to the state without PSS and the damping margin allowed relative to the original PSS parameter state, respectively. This represents the minimum damping threshold. By introducing... and It allows for small damping differences within a certain range caused by model errors, identification errors, or operational fluctuations, avoiding excessive sensitivity of the optimization process to minute damping changes; by introducing This ensures that the optimized PSS parameters will not fall below the minimum stability requirements under any operating conditions involved in the evaluation.
[0083] Subsequently, the damping ratio corresponding to the PSS parameter Θ to be optimized is further determined. With damping reference value The difference between them is used to construct the damping degradation amount under the i-th operating condition. The expression is:
[0084] when Greater than or equal to When the value is set to 'time', it indicates that the PSS parameters to be optimized have not caused damping degradation under this operating condition. The value is 0; when Less than When, it indicates that the PSS parameter to be optimized is lower than the damping requirement determined by the dual-benchmark constraints under this operating condition, and Di(Θ) takes the value of and The difference between them. Therefore, Penalties are only imposed on operating conditions with insufficient damping, and no additional penalties are imposed on operating conditions that already meet damping requirements.
[0085] Based on this, the damping degradation under each operating condition is weighted, squared, and summed to construct a dual-benchmark damping degradation constraint function. Its expression is:
[0086] in, The weight of the i-th operating condition is used to characterize the importance of this operating condition in the optimization of PSS parameters. For rated load or normal operating conditions, a normal weight can be set; for critical operating conditions with low stability margin or high engineering risk, such as deep peak shaving operating conditions, low load operating conditions, weak grid operating conditions, leading phase operating conditions, line interruption operating conditions, and peak shaving and frequency regulation superimposed operating conditions, a higher weight can be set so that the optimization algorithm prioritizes avoiding damping degradation under these critical operating conditions when searching for PSS parameters.
[0087] Furthermore, operating condition weights The weighting can be determined based on the probability of occurrence of the operating condition, the level of operational risk, the importance of dispatching, the size of the damping margin, or the statistical results of historical oscillation events. For example, the weighting can be increased for operating conditions that occur frequently in historical operations; similarly, the weighting can be increased for deep dispatching or weak grid operating conditions that, although less frequent, may cause low-frequency oscillation risks if they occur. In this way, the dual-benchmark damping degradation constraint can not only reflect mathematical damping insufficiency but also reflect the differences in engineering risks among different operating conditions.
[0088] It should be noted that the dual-benchmark damping degradation constraint constructed in this embodiment has a one-sided penalty characteristic. That is, it only applies when the PSS parameter to be optimized... The damping ratio under a certain operating condition The damping reference value is lower than that determined by the no-PSS state, the original PSS parameter state, and the minimum damping threshold. The constraint function only penalizes the parameter when the condition is met; if the parameter to be optimized reaches or exceeds the damping benchmark value under that operating condition, no damping degradation penalty is imposed. Therefore, even if a set of candidate PSS parameters can improve the average damping ratio under multiple operating conditions, as long as it causes a decrease in damping under some key deep peak-shaving conditions, weak grid conditions, or low-damping conditions, the candidate parameter will still be penalized. This avoids the problem of sacrificing the stability of key local operating conditions in exchange for improvements in overall damping performance.
[0089] S4. Based on the dual-benchmark damping degradation constraint, a comprehensive engineering constraint is constructed by combining at least one of the phase compensation over-constraint, modal frequency offset constraint, time-domain response availability constraint, parameter deviation constraint, and damping saturation constraint. A multi-condition comprehensive objective function is constructed and solved according to the comprehensive engineering constraint to obtain the power system stabilizer optimization parameters applicable to all operating conditions.
[0090] In this step, each constraint will be explained separately.
[0091] In one embodiment of the present invention, the PSS typically includes a lead-lag phase compensation stage. An over-constraint for phase compensation is used to limit the ratio of the time constant of the PSS compensation stage, the amplitude gain within the electromechanical oscillation frequency band, or the phase compensation deviation; the over-constraint for phase compensation is expressed as:
[0092] in, For electromechanical oscillations, the frequency band of concern is... The allowable amplitude magnification factor, For the phase of the PSS compensation channel under the i-th operating condition, To compensate for the phase of the target, To allow for phase deviation, L is the number of the lead-lag series. To optimize the frequency response of the phase compensation stage under the PSS parameters, This represents the frequency response of the phase compensation circuit under the original PSS parameters. This refers to the amplitude gain constraint in the phase compensation over-constraint. This is a time constant ratio constraint in the phase compensation over-constraint. This refers to the phase deviation constraint in the phase compensation over-constraint.
[0093] Furthermore, to avoid the optimization algorithm from causing an abnormal increase in the dominant electromechanical oscillation frequency in order to improve the damping ratio, this invention further constructs a modal frequency offset constraint to limit the offset of the dominant electromechanical modal frequency under the PSS parameters to be optimized relative to the original PSS parameter state. Specifically, let the dominant electromechanical modal frequency of the i-th operating condition under the original PSS parameters be . The dominant electromechanical mode frequency under the PSS parameters to be optimized is Then the modal frequency offset constraint for:
[0094] in, The dominant electromechanical mode frequency under the original power system stabilizer parameters for the i-th operating condition. The dominant electromechanical mode frequency under the power system stabilizer parameters to be optimized; To prevent the optimization algorithm from pursuing excessively high damping after the damping ratio has already met engineering requirements, leading to overly aggressive parameters, this invention also proposes a damping saturation constraint. This constraint reduces or stops rewarding increased damping once the damping index reaches the preset engineering requirements. Specifically, a damping saturation evaluation value is set:
[0095] in, This represents the upper limit of the damping reward. Once the damping ratio exceeds this upper limit, the objective function will no longer reward its increase, thereby improving the conservatism of the optimization results and the engineering usability.
[0096] In addition, this application proposes parameter deviation constraints to limit the degree of deviation of the PSS parameters to be optimized from the original PSS parameters.
[0097] Because there may be differences between the small-signal model and the electromagnetic transient model, such as excitation limiting, sampling delay, filtering implementation, PSS output limiting, speed signal calculation method, and shaft system model, this invention further introduces time-domain response availability constraints to evaluate the engineering availability of PSS parameters based on the active power initial swing amplitude, oscillation convergence time, PSS output amplitude, PSS output change rate, terminal voltage deviation, or number of oscillations. For the i-th operating condition, the time-domain response index is defined as follows: For the i-th operating condition, define the time-domain response index. :
[0098] Furthermore, the time-domain response index described in this invention is used to construct a time-domain response penalty term based on a preset acceptable threshold. , is represented as:
[0099] in, The active power initial swing amplitude, To shorten the convergence time, Output the maximum value for PSS. For PSS output limits, The number of oscillations. The acceptable threshold for time-domain response is defined. In the first stage of optimization, candidate parameters can be obtained using only small-signal damping constraints. In the second stage of screening, parameter combinations with good small-signal performance but poor time-domain response are eliminated through electromagnetic transient simulation or field disturbance response.
[0100] Finally, considering the above indicators, this invention establishes the following multi-condition integrated objective function:
[0101]
[0102]
[0103]
[0104] in, This is a penalty term for modal frequency shift. This is a phase deviation penalty term. As a time-domain response availability penalty, This is a penalty term for parameters deviating from their original values. , , to The weighting coefficient is used; the time-domain response availability penalty is constructed based on at least two of the following indicators: active power initial swing amplitude, oscillation convergence time, PSS output maximum amplitude, PSS output rate of change, maximum deviation of terminal voltage, and number of oscillations.
[0105] It is understood that this application does not specifically limit the optimization algorithm, and it can be particle swarm optimization, genetic algorithm, differential evolution algorithm, Bayesian optimization algorithm, sequential quadratic programming algorithm, or other algorithms capable of solving nonlinear constrained optimization problems.
[0106] Furthermore, to avoid negative damping caused by differences in PSS output polarity, power channel symbol, per-unit reference, or module implementation between the small-signal model, electromagnetic transient model, and actual excitation device, this invention further performs a self-verification of PSS injection polarity and parameter mapping before or after PSS parameter optimization. Specifically, Define the polarity variable σ∈{PSS output injection polarity variable} 1, +1}, calculate the multi-condition damping level corresponding to the original PSS parameters under the two polarities respectively, and determine:
[0107] Where σ is the PSS output injection polarity variable. When the difference in minimum damping ratio between the two polarities exceeds a set threshold, σ will be... As the polarity verification result of PSS model parameter mapping and engineering download; when the difference does not exceed the set threshold, a polarity uncertainty prompt is output, and further verification is performed through electromagnetic transient small disturbance test or field small disturbance test.
[0108] In one exemplary embodiment, operating conditions of a thermal power unit under active power outputs of 0.95 pu, 0.35 pu, 0.25 pu, and 0.10 pu are selected. Small-signal models of the synchronous generator, excitation system, PSS, and external power grid are established respectively. The damping ratio and mode frequency of the dominant electromechanical modes are calculated under the conditions of no PSS, original PSS parameters, and PSS parameters to be optimized. Based on this, comprehensive engineering constraints are constructed and key PSS parameters are solved. Subsequently, the active power first swing amplitude, oscillation convergence time, PSS output amplitude, and generator terminal voltage deviation are checked in electromagnetic transient simulation. Parameter combinations with good small-signal damping indicators but unacceptable time-domain response are screened out. Finally, PSS parameters that meet the requirements of full-condition damping improvement and engineering usability are output.
[0109] Example 2 This invention also provides a full-condition optimization system 10 for PSS parameters of thermal power units, such as... Figure 2 As shown, the system includes: The multi-condition generation module 100 is used to construct a set of multiple operating conditions covering thermal power units under different active power output, reactive power output, terminal voltage, external power grid strength and operating mode. The damping evaluation module 200 is used to establish a linearized model of the synchronous generator, excitation system, power system stabilizer (PSS) and external power grid for each operating condition in the set of multiple operating conditions to obtain the dominant electromechanical oscillation mode, or to obtain the dominant electromechanical oscillation mode through response identification, so as to establish a damping evaluation model under multiple operating conditions. The constraint construction module 300 is used to construct a damping evaluation set that includes the state without a power system stabilizer, the original power system stabilizer parameter state, and the power system stabilizer parameter state to be optimized. Based on the damping evaluation set, a dual-benchmark damping degradation constraint is constructed to avoid damping degradation of the parameters to be optimized under certain operating conditions during the optimization process. The parameter optimization and output module 400 is used to construct comprehensive engineering constraints based on the dual-benchmark damping degradation constraints, combined with at least one of phase compensation over-constraint, modal frequency offset constraints, time-domain response availability constraints, parameter deviation constraints, and damping saturation constraints. Based on the comprehensive engineering constraints, a multi-condition comprehensive objective function is constructed and solved to obtain power system stabilizer optimization parameters applicable to all operating conditions.
[0110] Regarding the system in the above embodiments, the specific ways in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated here.
[0111] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
[0112] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.
[0113] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.
Claims
1. A method for optimizing the PSS parameters of a thermal power unit under all operating conditions, characterized in that, Includes the following steps: S1, construct a set of multiple operating conditions covering thermal power units under different active power output, reactive power output, terminal voltage, external power grid strength and operating mode; S2, for each operating condition in the set of multiple operating conditions, establish a linearized model of the synchronous generator, excitation system, power system stabilizer (PSS) and external power grid to obtain the dominant electromechanical oscillation mode, or obtain the dominant electromechanical oscillation mode through response identification, so as to establish a damping evaluation model under multiple operating conditions. S3. Construct a damping evaluation set that includes the state without a power system stabilizer, the original power system stabilizer parameter state, and the power system stabilizer parameter state to be optimized. Based on the damping evaluation set, construct a dual-benchmark damping degradation constraint to avoid damping degradation of the parameters to be optimized under certain operating conditions during the optimization process. S4. Based on the dual-benchmark damping degradation constraint, a comprehensive engineering constraint is constructed by combining at least one of the phase compensation over-constraint, modal frequency offset constraint, time-domain response availability constraint, parameter deviation constraint, and damping saturation constraint. A multi-condition comprehensive objective function is constructed and solved according to the comprehensive engineering constraint to obtain the power system stabilizer optimization parameters applicable to all operating conditions.
2. The method according to claim 1, characterized in that, The process of constructing a set of multiple operating conditions includes: The system acquires operating data of thermal power units under at least one of the following operating conditions: rated or high load operation, low load operation, deep peak shaving operation, leading phase operation, operation under different external power grid strengths, and operation under different reactive power outputs. Based on this operating data, a set of multiple operating conditions is constructed, wherein the set of multiple operating conditions is represented as follows: in, For the active power output of the unit under the i-th operating condition, To exert effort for no reason This refers to the terminal voltage or bus voltage. Equivalent impedance to the external power grid. This is the operating mode identifier, and N is the number of operating conditions.
3. The method according to claim 1, characterized in that, For each operating condition in the set of multiple operating conditions, a linearized model of the synchronous generator, excitation system, power system stabilizer, and external power grid is established to obtain the dominant electromechanical oscillation mode, or the dominant electromechanical oscillation mode is obtained through response identification, including: For each operating condition, the dominant electromechanical oscillation mode parameters are obtained through at least one of the following methods: small-signal modal analysis, simulation response identification, field disturbance response identification, or online measurement data identification. These dominant electromechanical oscillation mode parameters include at least the damping ratio and the modal frequency.
4. The method according to claim 1, characterized in that, The construction includes a damping evaluation set comprising the state without a power system stabilizer, the original power system stabilizer parameter state, and the power system stabilizer parameter state to be optimized. Based on this damping evaluation set, a dual-benchmark damping degradation constraint is constructed, including: For each operating condition, calculate the dominant electromechanical mode damping ratio when the power system stabilizer is not engaged. Dominant electromechanical mode damping ratio when using original power system stabilizer parameters and the dominant electromechanical mode damping ratio when using the parameters of the power system stabilizer to be optimized. ; The key PSS parameter that significantly affects the damping of low-frequency oscillations was selected as the optimization variable, namely... The remaining parameters are fixed according to the original setting value of the excitation device or the typical engineering value; The dual-benchmark damping degradation constraint is constructed as follows: The formula is: in, The dominant electromechanical mode damping ratio for the i-th operating condition without PSS. The dominant electromechanical mode damping ratio for the i-th operating condition under the original PSS parameter state. To determine the dominant electromechanical mode damping ratio for the i-th operating condition under the PSS parameter Θ to be optimized, and To allow for damping margin, The minimum damping threshold, Let be the weight of the i-th operating condition.
5. The method according to claim 1, characterized in that, Phase compensation over-constraint is used to limit the ratio of time constants in the PSS compensation stage, the amplitude gain within the electromechanical oscillation frequency band, or the phase compensation deviation; the phase compensation over-constraint is expressed as: in, For electromechanical oscillations, the frequency band of concern is... The allowable amplitude magnification factor, For the phase of the PSS compensation channel under the i-th operating condition, To compensate for the phase of the target, To allow for phase deviation, L is the number of the lead-lag series. To optimize the frequency response of the phase compensation circuit under the PSS parameters, This represents the frequency response of the phase compensation circuit under the original PSS parameters. This refers to the amplitude gain constraint in the phase compensation over-constraint. This is a time constant ratio constraint in the phase compensation over-constraint. This refers to the phase deviation constraint in the phase compensation over-constraint.
6. The method according to claim 1, characterized in that, Modal frequency shift constraints are used to limit the shift of the dominant electromechanical mode frequency relative to the original PSS parameter state under the PSS parameters to be optimized. The process of constructing the modal frequency shift constraints includes: Let the dominant electromechanical mode frequency of the i-th operating condition under the original PSS parameters be... The dominant electromechanical mode frequency under the PSS parameters to be optimized is Then the modal frequency offset constraint for: in, The dominant electromechanical mode frequency under the original power system stabilizer parameters for the i-th operating condition. The dominant electromechanical mode frequency under the power system stabilizer parameters to be optimized; Parameter deviation constraints are used to limit the degree of deviation of the PSS parameters to be optimized from the original PSS parameters; Damping saturation constraint is used to reduce or stop further damping increases after the damping index reaches the preset engineering requirements. The damping saturation evaluation value is set as follows: ,in This is the upper limit of the damping reward.
7. The method according to claim 1, characterized in that, Time-domain response availability constraints are used to evaluate the engineering availability of PSS parameters based on the active power first swing amplitude, oscillation convergence time, PSS output amplitude, PSS output rate of change, terminal voltage deviation, or number of oscillations. The process of constructing time-domain response availability constraints includes: For the i-th operating condition, define the time-domain response index. : A time-domain response penalty is constructed based on the time-domain response index and a preset acceptable threshold. , is represented as: in, The active power initial swing amplitude, To shorten the convergence time, Output the maximum value for PSS. For PSS output limits, The number of oscillations. This represents the acceptable threshold for the time-domain response.
8. The method according to claim 7, characterized in that, The multi-condition integrated objective function is expressed as follows: in, This is a penalty term for modal frequency shift. This is a phase deviation penalty term. As a time-domain response availability penalty, This is a penalty term for parameters deviating from their original values. , , to The weighting coefficient is used; the time-domain response availability penalty is constructed based on at least two of the following indicators: active power initial swing amplitude, oscillation convergence time, PSS output maximum amplitude, PSS output rate of change, maximum deviation of terminal voltage, and number of oscillations.
9. The method according to claim 8, characterized in that, Also includes: Before or after PSS parameter optimization, perform a self-verification of PSS injection polarity and parameter mapping. The self-verification steps include: Define the polarity variable σ∈{PSS output injection polarity variable} 1, +1}, calculate the multi-condition damping level corresponding to the original PSS parameters under the two polarities respectively, and determine: Where σ is the polarity variable injected into the PSS output; When the difference in minimum damping ratio between the two polarities exceeds a set threshold, σ As the polarity verification result of PSS model parameter mapping and engineering download; when the difference does not exceed the set threshold, a polarity uncertainty prompt is output, and further verification is performed through electromagnetic transient small disturbance test or field small disturbance test.
10. A full-condition optimization system for PSS parameters of thermal power units, characterized in that, include: The multi-condition generation module is used to construct a set of multiple operating conditions covering thermal power units under different active power output, reactive power output, terminal voltage, external power grid strength and operating modes. The damping evaluation module is used to establish a linearized model of the synchronous generator, excitation system, power system stabilizer (PSS), and external power grid for each operating condition in the set of multiple operating conditions to obtain the dominant electromechanical oscillation mode, or to obtain the dominant electromechanical oscillation mode through response identification, so as to establish a damping evaluation model under multiple operating conditions. The constraint construction module is used to construct a damping evaluation set that includes the state without a power system stabilizer, the original power system stabilizer parameter state, and the power system stabilizer parameter state to be optimized. Based on the damping evaluation set, a dual-benchmark damping degradation constraint is constructed to avoid damping degradation of the parameters to be optimized under certain operating conditions during the optimization process. The parameter optimization and output module is used to construct comprehensive engineering constraints based on the dual-benchmark damping degradation constraints, combined with at least one of phase compensation over-constraint, modal frequency offset constraints, time-domain response availability constraints, parameter deviation constraints, and damping saturation constraints. Based on the comprehensive engineering constraints, a multi-condition comprehensive objective function is constructed and solved to obtain power system stabilizer optimization parameters applicable to all operating conditions.
