A test system for mismatch correction of speed regulation model of thermal power units with wide load range

Abstract

The present invention relates to the field of steam turbine testing technology, and in particular to a test system for correcting mismatch in speed regulation models of thermal power units in a wide load range. The test system comprises a thermal system component, a speed regulation system device, a dynamic signal acquisition device and a control device. A dynamic signal processing module is used to analyze in real time the correlation between sudden changes in main steam pressure and throttle opening oscillations, and a dynamic response database including time series is constructed. A throttle opening feedforward compensation instruction and a burner fuel distribution gradient adjustment signal are generated according to correction parameters, thereby suppressing power overregulation peaks caused by sudden load changes and reducing the delay of primary frequency regulation actions. A simulation verification module compares measured data with model outputs through an error feedback mechanism, dynamically adjusts calibration parameters of a high-pressure cylinder flow characteristic curve, and optimizes a control strategy in combination with a grid-related safety margin assessment result, thereby improving the dynamic response accuracy of the entire peak regulation range and enhancing the frequency stability of the power grid.

Description

Technical Field

[0001] The invention relates to the technical field of steam turbine testing, in particular to a testing system for correcting mismatch of a speed regulation model of a thermal power unit in a wide load range. Background Art

[0002] The overshoot peak of the turbine power curve refers to the transient fluctuation phenomenon in which the output power exceeds the target set value due to sudden load changes or control lag during operation. It belongs to the research scope of dynamic characteristics of power machinery. The mechanical vibration, speed changes and strain response of key components corresponding to the overshoot peak are captured through dynamic signal acquisition devices and spectrum analysis technology. The fluctuation amplitude and attenuation rate are quantified by combining time series models, and then the control algorithm parameters are optimized or the damping structure design is improved to suppress the impact of overshoot behavior on system stability under non-steady-state conditions.

[0003] Steam turbines in thermal power plants face challenges during deep peak regulation, including poor coordinated control quality, degraded primary frequency regulation performance, and unstable combustion. This is primarily due to significant fluctuations in key parameters such as main steam pressure, temperature, and heating extraction steam flow across a wide range of load conditions, which prevents the speed regulation system model from accurately characterizing its dynamic response characteristics. Furthermore, existing testing methods are limited to calibration under a single operating condition and fail to cover the dynamic processes of the entire peak regulation range. For example, sudden changes in main steam pressure during low-load operation cause frequent oscillations in the throttle valve opening, leading to unstable combustion and delayed primary frequency regulation, directly impacting grid frequency stability. Furthermore, the nonlinear variations in the high-throttle valve flow characteristics of steam turbines across a wide load range have not been accurately modeled, resulting in significant deviations between simulation results and measured responses, making it impossible to effectively assess the unit's grid safety margin. Summary of the Invention

[0004] In response to the shortcomings of the existing technology, the present invention provides a test system for correcting the mismatch of the speed control model of thermal power units in a wide load range, which solves the problem of speed control system performance test deviation caused by dynamic response mismatch of the steam turbine of the thermal power unit during deep peak regulation in the existing technology, which is specifically manifested in inaccurate controller response characteristics, decreased speed regulation accuracy and dynamic fluctuations of combustion parameters.

[0005] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows:

[0006] The present invention provides a test system for correcting mismatch of speed regulation model of thermal power unit in wide load range, comprising: a thermal system component, a speed regulation system device, a dynamic signal acquisition device and a control device, wherein the control device establishes communication connection with the thermal system component, the speed regulation system device and the dynamic signal acquisition device;

[0007] The dynamic signal acquisition device includes a high-precision sensor that collects main steam pressure, temperature, valve opening, speed deviation signal and energy storage chamber temperature gradient data in real time, and transmits the signal to the control device;

[0008] The control device includes a dynamic signal processing module, a parameter identification module, a closed-loop control module and a simulation verification module;

[0009] The dynamic signal processing module receives the sensor signal, analyzes the main steam parameter change trend, the valve oscillation characteristics and the phase change material heat capacity attenuation characteristics, generates a dynamic response database including a time series, and transmits the dynamic response database to the parameter identification module;

[0010] The parameter identification module identifies the regulator response hysteresis parameters, actuator nonlinear characteristic parameters, prime mover dynamic response parameters, and phase change material heat capacity attenuation coefficient in the speed control system device through a multi-objective rolling optimization algorithm based on the dynamic disturbance test data in the dynamic response database, corrects the model mismatch of the preset speed control system model under a wide load condition, and obtains the corrected parameters to generate a throttle opening instruction and a combustion control signal;

[0011] The closed-loop control module generates a throttle opening instruction and a combustion control signal based on the corrected parameters, and drives the speed control system device to perform turbine power adjustment and burner fuel distribution. The simulation verification module receives real-time data from the dynamic signal acquisition device and the corrected model output of the parameter identification module, compares the measured response with the simulation result through the error feedback mechanism, dynamically adjusts the calibration parameters of the high-pressure cylinder flow characteristic curve, calculates the power fluctuation allowable threshold, the frequency modulation response time margin and the frequency deviation safety range based on the comparison error data and the fouling compensation model verification result, generates a grid-related safety margin assessment result, and feeds the assessment result back to the closed-loop control module to optimize the constraint conditions of the throttle opening instruction and the combustion control signal.

[0012] Furthermore, in the test system for mismatch correction of the speed regulation model of a thermal power unit with a wide load range described in the present invention, the dynamic signal processing module analyzes the nonlinear fluctuation characteristics in the sensor signal transmitted by the dynamic signal acquisition device, synchronously integrates the energy storage chamber temperature gradient data and the molten salt scaling factor change rate, identifies the dynamic correlation between the sudden change of main steam pressure, the throttle opening oscillation and the phase change material heat capacity attenuation, and generates the dynamic response database including the time series of the main steam pressure change rate, the throttle opening fluctuation amplitude, the speed deviation, the high-temperature phase change material heat capacity attenuation slope and the molten salt scaling factor dynamic offset;

[0013] The heat capacity attenuation slope of the high-temperature phase change material is calculated by the following formula:

[0014] α=-(ΔT gradient / ΔtΔQ heat capacity);

[0015] ΔQ heat capacity is the heat capacity loss per unit time of the energy storage cavity, ΔT gradient is the temperature difference between the high / medium temperature energy storage cavity, and Δt is the sampling period;

[0016] The correlation data in the dynamic response database is input into the parameter identification module to correct the regulator response hysteresis parameters and the phase change material thermodynamic performance and scaling coupling characteristic parameters in the preset speed control system model.

[0017] Furthermore, in the test system for mismatch correction of speed regulation model of thermal power units in a wide load range according to the present invention, the parameter identification module is configured as follows:

[0018] Obtaining the limiting parameters, dead zone parameters, and initial heat capacity attenuation coefficient of the phase change material of the energy storage cavity in the speed regulation system device through static tests, and inputting the parameters into the speed regulation system model;

[0019] Based on the dynamic disturbance test, the input and output signals of the regulator, the input and output signals of the actuator, the input and output signals of the prime mover, the temperature gradient data of the high-temperature or medium-temperature energy storage chamber, and the real-time offset of the molten salt scaling factor in the speed control system device are collected;

[0020] A multi-objective rolling optimization algorithm is used to iteratively calculate the input and output signals of the regulator, the input and output signals of the actuator, the input and output signals of the prime mover, and the energy storage and scaling coupling parameters, so as to identify the following in the speed control system model: the regulator response hysteresis parameter, the actuator nonlinear characteristic parameter, the dynamic compensation coefficient of the phase change material heat capacity attenuation, and the coupling weight of the scaling factor and the heat capacity attenuation;

[0021] The objective function of the multi-objective rolling optimization algorithm is:

[0022] min(w1*E peak load regulation+w2*R power abandonment+w3*D thermal attenuation);

[0023] E peak regulation is the load tracking error, R curtailment is the wind and solar curtailment rate, and D thermal attenuation is the thermal capacity attenuation deviation; weights w1=0.4, w2=0.35, w3=0.25;

[0024] The corrected parameters are transmitted to the closed-loop control module.

[0025] Furthermore, in the test system for correcting mismatch of speed regulation model of thermal power unit with wide load range described in the present invention, the closed-loop control module generates a throttle opening instruction to adjust the power output of the steam turbine based on the corrected regulator response hysteresis parameter and actuator nonlinear characteristic parameter transmitted by the parameter identification module, thereby suppressing overregulation peak caused by sudden load change;

[0026] At the same time, based on the dynamic response parameters of the prime mover corrected by the parameter identification module, a combustion control signal is generated to adjust the fuel distribution ratio of the burner.

[0027] Furthermore, in the test system for correcting mismatch of speed control model of thermal power unit in wide load range described in the present invention, the simulation verification module receives the measured throttle oscillation frequency and main steam pressure attenuation rate generated by the dynamic signal processing module, and compares them with the dynamic response output by the simulation model corrected by the parameter identification module, adjusts the characteristic curve calibration parameters in the speed control system model through the error feedback mechanism, and transmits the corrected grid-related safety margin assessment result to the closed-loop control module for optimizing the generation logic of the throttle opening instruction and the combustion control signal.

[0028] Furthermore, in the test system for mismatch correction of speed regulation model of thermal power units in a wide load range according to the present invention, the dynamic disturbance test includes:

[0029] Injecting a step disturbance signal into the speed control system device within a wide load range of 100% rated load deep peak regulation lower limit, and collecting the throttle valve opening oscillation waveform output by the actuator when the main steam pressure suddenly changes;

[0030] The frequency characteristics of the throttle opening oscillation waveform are extracted through spectrum analysis, and the frequency characteristics are input into the speed control system model. The dynamic response parameters of the prime mover corrected by the parameter identification module are checked, and the corrected prime mover model parameters are generated and transmitted to the closed-loop control module.

[0031] Furthermore, in the test system for correcting mismatch of speed regulation model of thermal power units in a wide load range described in the present invention, the dynamic signal processing module performs spectrum correction processing on the throttle opening signal transmitted by the dynamic signal acquisition device, eliminates high-frequency noise interference introduced in the sensor acquisition process through digital filtering, generates a corrected throttle opening signal, and inputs the corrected throttle opening signal into the closed-loop control module, which is used to generate a throttle opening compensation instruction based on the regulator response lag parameter corrected by the parameter identification module, and adjust the turbine throttle opening.

[0032] Furthermore, the test system for correcting mismatch of a speed regulation model of a thermal power unit in a wide load range according to the present invention further includes:

[0033] The speed regulation system device is connected to a preset two-stage energy storage structure, which includes: a low-pressure energy storage chamber, a medium-pressure energy storage chamber and a high-pressure energy storage chamber;

[0034] The high-pressure energy storage chamber and the medium-pressure energy storage chamber are filled with high-temperature phase change materials, and the melting point of the high-temperature phase change materials is ≥300℃;

[0035] Low-pressure energy storage chamber, filled with medium-temperature phase change material, 200℃≤medium-temperature phase change material melting point<300℃;

[0036] The speed control system device includes a hydraulic regulator, a PID controller and an actuator, which is used to dynamically adjust the opening of the turbine valve to control the power output.

[0037] Furthermore, the test system for correcting mismatch of a speed regulation model of a thermal power unit in a wide load range according to the present invention further includes:

[0038] The multi-objective rolling optimization algorithm simultaneously optimizes peak shaving depth, equipment life, and wind and solar curtailment rates;

[0039] Dynamic binding of renewable energy consumption priorities to energy storage and release strategies;

[0040] and transmitting the corrected model parameters to the closed-loop control module;

[0041] The closed-loop control module generates a throttle opening instruction and a combustion control signal according to the corrected model parameters, drives the actuator to adjust the turbine power output, suppresses the over-modulation peak caused by the sudden change in load, and optimizes the burner fuel distribution to reduce the delay of the primary frequency modulation action.

[0042] Furthermore, the test system for correcting mismatch of a speed regulation model of a thermal power unit in a wide load range according to the present invention further includes:

[0043] The simulation verification module compares the measured data generated by the dynamic signal processing module with the simulation model output corrected by the parameter identification module, dynamically adjusts the model characteristic curve calibration parameters through the error feedback mechanism, and verifies the real-time compensation model for the scaling factor of the thermoelectric decoupling heat exchanger:

[0044] δt=δ0×e-kt+β×ΔTmolten salt;

[0045] Where δ0 is the initial scaling factor, k is the attenuation coefficient, β is the temperature compensation coefficient, and ΔTmoltensalt is the molten salt temperature difference.

[0046] Beneficial effects of the present invention:

[0047] The present invention uses a dynamic signal processing module to analyze the correlation characteristics between sudden changes in main steam pressure and throttle opening oscillations in real time, and constructs a dynamic response database including time series; the parameter identification module adopts a multi-objective rolling optimization algorithm based on the database to identify the regulator response lag parameters, actuator nonlinear characteristic parameters and prime mover dynamic response parameters in the speed control system, and correct the model mismatch under wide load conditions; the closed-loop control module generates a throttle opening feedforward compensation instruction and a burner fuel distribution gradient adjustment signal according to the corrected parameters, suppresses the power overmodulation peak caused by load mutation through a phase advance compensation mechanism, and optimizes the fuel distribution ratio to reduce the primary frequency modulation action delay; the simulation verification module dynamically adjusts the calibration parameters of the high-pressure cylinder flow characteristic curve through an error feedback mechanism, and optimizes the control strategy in combination with the grid-related safety margin assessment results, ultimately achieving improved dynamic response accuracy in the entire peak-shaving interval, enhanced combustion stability and optimized grid frequency regulation capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, for ordinary technicians in this field, other drawings can be obtained based on the drawings without paying any creative labor.

[0049] Figure 1 A system architecture diagram of a test system for correcting mismatches in speed regulation models of thermal power units over a wide load range, provided by an embodiment of the present invention.

[0050] Figure 2 A dynamic signal processing flow chart for correcting over-modulation peaks in a steam turbine power curve provided by an embodiment of the present invention.

[0051] Figure 3 A parameter identification flow chart for correcting overshoot peaks in a steam turbine power curve provided by an embodiment of the present invention.

[0052] Figure 4 A closed-loop control logic block diagram for correcting over-regulated peaks in a steam turbine power curve provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the specific embodiments of the present invention and the corresponding drawings. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. The technical solutions provided by each embodiment of the present invention are described in detail below in conjunction with the drawings. In order to better understand the purpose of the present invention, the present invention is further described in detail below.

[0054] See also Figures 1 to 4 The present invention provides a test system for correcting mismatch of speed regulation model of thermal power unit in wide load range, comprising: a thermal system component, a speed regulation system device, a dynamic signal acquisition device and a control device, wherein the control device establishes communication connection with the thermal system component, the speed regulation system device and the dynamic signal acquisition device;

[0055] The dynamic signal acquisition device includes a high-precision sensor that collects main steam pressure, temperature, valve opening, speed deviation signal and energy storage chamber temperature gradient data in real time, and transmits the signal to the control device;

[0056] The control device includes a dynamic signal processing module, a parameter identification module, a closed-loop control module and a simulation verification module;

[0057] The dynamic signal processing module receives the sensor signal, analyzes the main steam parameter change trend, the valve oscillation characteristics and the phase change material heat capacity attenuation characteristics, generates a dynamic response database including a time series, and transmits the dynamic response database to the parameter identification module;

[0058] The parameter identification module identifies the regulator response hysteresis parameters, actuator nonlinear characteristic parameters, prime mover dynamic response parameters, and phase change material heat capacity attenuation coefficient in the speed control system device through a multi-objective rolling optimization algorithm based on the dynamic disturbance test data in the dynamic response database, corrects the model mismatch of the preset speed control system model under a wide load condition, and obtains the corrected parameters to generate a throttle opening instruction and a combustion control signal;

[0059] The closed-loop control module generates a throttle opening instruction and a combustion control signal based on the corrected parameters, and drives the speed control system device to perform turbine power adjustment and burner fuel distribution. The simulation verification module receives real-time data from the dynamic signal acquisition device and the corrected model output of the parameter identification module, compares the measured response with the simulation result through the error feedback mechanism, dynamically adjusts the calibration parameters of the high-pressure cylinder flow characteristic curve, calculates the power fluctuation allowable threshold, the frequency modulation response time margin and the frequency deviation safety range based on the comparison error data and the fouling compensation model verification result, generates a grid-related safety margin assessment result, and feeds the assessment result back to the closed-loop control module to optimize the constraint conditions of the throttle opening instruction and the combustion control signal.

[0060] The thermal system components simulate the actual thermal cycle of a thermal power unit. The main steam parameters generated by the boiler are collected in real time by high-precision pressure and temperature sensors with a sampling frequency of no less than 1kHz, covering the entire peak-shaving operating range from rated load to the lower limit of deep peak-shaving. The speed regulation system integrates a hydraulic actuator with a two-stage energy storage structure. The high- and medium-pressure energy storage chambers are filled with high-temperature phase-change materials with a melting point of 300°C or higher, while the low-pressure energy storage chamber is filled with medium-temperature phase-change materials with a melting point range of 200-300°C. The dynamic signal acquisition device simultaneously collects valve opening displacement signals, speed deviation signals, and multi-point temperature gradient data in the energy storage chamber. The molten salt scaling factor is indirectly measured using electrochemical impedance spectroscopy. All sensor data is converted into standardized electrical signals by signal conditioning circuits.

[0061] The dynamic signal processing module preprocesses the raw signal: a Butterworth low-pass filter is used to eliminate high-frequency noise interference while retaining the fundamental frequency component of the throttle valve oscillation. A time window integration algorithm is used to capture the instantaneous rate of change of the main steam pressure, and a sliding window is used to calculate the amplitude of the throttle valve opening fluctuation. Furthermore, the thermal capacity decay slope of the phase change material is calculated by integrating the temperature difference data of the energy storage chamber, generating a timestamp-aligned dynamic response database. This database contains four key feature vectors: the rate of change of the main steam pressure, the amplitude of the throttle valve opening fluctuation, the thermal capacity decay slope, and the offset of the molten salt scaling factor. The data structure is hierarchical and stores the raw sample values, analytical features, and a correlation matrix.

[0062] The parameter identification module performs both static and dynamic disturbance tests. The static test determines the actuator displacement deadband threshold and limit range. The dynamic disturbance test injects a standardized step disturbance signal across a wide load range, synchronously collecting the regulator command input and displacement feedback output signals. A multi-objective rolling optimization algorithm uses load tracking error, wind and solar curtailment rate, and thermal capacity decay deviation as comprehensive targets. It solves the problem in a piecewise iterative manner using preset weights. It outputs four physically measurable parameters: the regulator response lag time constant, the actuator nonlinear friction coefficient, the thermal capacity decay dynamic compensation coefficient, and the scaling coupling weight. These parameters are then used to correct the nonlinear gain coefficient of the high-pressure cylinder flow characteristic curve in the speed control system model.

[0063] The closed-loop control module converts the correction parameters into executable instructions. A phase advance transfer function is constructed based on the regulator lag time to generate a throttle valve opening feedforward compensation. The turbine throttle valve opening position is adjusted via the oil motor servo drive mechanism. The fuel distribution gradient is calculated based on the prime mover's volumetric time constant, and the fuel supply weights for high and low temperature burners are optimized based on the accumulator chamber temperature distribution. Actuator displacement feedback signals and burner status monitoring data are transmitted in real time to the dynamic signal acquisition device to optimize the instructions for the next control cycle.

[0064] The simulation verification module compares the real-time throttle oscillation waveform of the dynamic signal acquisition device with the corrected model output of the parameter identification module, uses the dynamic time warping algorithm to align the time domain curve, and calculates the phase lag and amplitude deviation residuals. The residual results are iteratively adjusted using the gradient descent method to calibrate the high-pressure cylinder flow characteristic calibration parameters, while verifying the effectiveness of the attenuation coefficient and temperature compensation coefficient in the fouling compensation model. Based on the residual analysis results and fouling verification data, three quantitative indicators are calculated: the power fluctuation allowable threshold, the frequency modulation response time margin, and the frequency deviation safety range, to generate a grid-related safety margin assessment result. The assessment result is converted into a throttle opening change rate limit and a fuel distribution gradient constraint, which are fed back to the closed-loop control module to reconstruct the instruction generation logic.

[0065] Specifically, the test system for correcting mismatch in the speed regulation model of a thermal power unit with a wide load range described in the present invention comprises a dynamic signal processing module that analyzes the nonlinear fluctuation characteristics in the sensor signal transmitted by the dynamic signal acquisition device, synchronously integrates the energy storage chamber temperature gradient data and the molten salt scaling factor change rate, identifies the dynamic correlation between the main steam pressure sudden change, the throttle opening oscillation, and the phase change material heat capacity attenuation, and generates the dynamic response database including the time series of the main steam pressure change rate, the throttle opening fluctuation amplitude, the speed deviation, the high-temperature phase change material heat capacity attenuation slope, and the molten salt scaling factor dynamic offset;

[0066] The heat capacity attenuation slope of the high-temperature phase change material is calculated by the following formula:

[0067] α=-(ΔT gradient / ΔtΔQ heat capacity);

[0068] ΔQ heat capacity is the heat capacity loss per unit time of the energy storage cavity, ΔT gradient is the temperature difference between the high / medium temperature energy storage cavity, and Δt is the sampling period;

[0069] The correlation data in the dynamic response database is input into the parameter identification module to correct the regulator response hysteresis parameters and the phase change material thermodynamic performance and scaling coupling characteristic parameters in the preset speed control system model.

[0070] The dynamic signal processing module receives multiple sensor signals from the dynamic signal acquisition device, including data on main steam pressure, temperature, throttle valve opening, speed deviation, and accumulator temperature gradient. This module first preprocesses the raw signal, using a digital filtering algorithm to eliminate high-frequency noise and improve the signal-to-noise ratio. After preprocessing, the signal enters the feature analysis phase, extracting the instantaneous rate of change of the main steam pressure through time-domain analysis. A sliding window algorithm is also used to capture the oscillation amplitude of the throttle valve opening signal.

[0071] After extracting basic features, the module simultaneously fuses the energy storage chamber temperature gradient data with the molten salt scaling factor change rate information. This data is collected using a multi-point temperature sensor array, reflecting the heat transfer state of the high- and medium-temperature phase change materials. The molten salt scaling factor change rate is indirectly measured using electrochemical impedance spectroscopy, characterizing the dynamic accumulation of deposits on the phase change material surface. A timestamp alignment mechanism is used during the fusion process to ensure the synchronization of multi-source data.

[0072] Based on fused data, this module identifies the dynamic correlations between main steam pressure fluctuations, throttle opening oscillations, and phase change material heat capacity decay. Sudden main steam pressure fluctuations are characterized by a pressure change rate exceeding a set threshold. The throttle opening oscillations are characterized by extracting the amplitude of the fundamental frequency component through spectral analysis. The phase change material heat capacity decay slope is calculated based on the ratio of temperature difference to heat capacity loss. The correlations between these three factors are quantified using a correlation coefficient matrix, establishing a coupled pressure-oscillation-attenuation model.

[0073] This module generates a dynamic response database consisting of time series. The data structure includes the main steam pressure change rate, throttle opening fluctuation amplitude, speed deviation, high-temperature phase change material heat capacity decay slope, and molten salt scaling factor dynamic offset. The main steam pressure change rate records the pressure change per unit time, the throttle opening fluctuation amplitude calculates the maximum displacement deviation within the oscillation cycle, the heat capacity decay slope indicates the degree of phase change material heat capacity loss per unit time, and the molten salt scaling factor dynamic offset reflects the coefficient of influence of the scaling layer thickness on heat transfer efficiency.

[0074] The dynamic response database uses a layered storage architecture: the base layer stores raw sample values, the feature layer stores parsed parameter indicators, and the correlation layer stores multi-parameter coupling matrices. The database output is standardized to unify dimensions and sampling frequencies, and mappings between parameters are established using timestamp indexes.

[0075] After receiving correlation data from the dynamic response database, the parameter identification module focuses on correcting the regulator response hysteresis parameters and the phase change material thermodynamic performance and fouling coupling characteristic parameters in the speed control system model. The regulator response hysteresis parameter correction is based on a phase difference analysis between the main steam pressure change rate and the throttle action delay. The fouling coupling characteristic parameter correction is based on a regression model of the heat capacity decay slope and the fouling factor offset. These corrected parameters are input into the speed control system model to optimize the dynamic response accuracy under variable operating conditions.

[0076] Specifically, in the test system for mismatch correction of speed regulation model of thermal power units in a wide load range according to the present invention, the parameter identification module is configured as follows:

[0077] Obtaining the limiting parameters, dead zone parameters, and initial heat capacity attenuation coefficient of the phase change material of the energy storage cavity in the speed regulation system device through static tests, and inputting the parameters into the speed regulation system model;

[0078] Based on the dynamic disturbance test, the input and output signals of the regulator, the input and output signals of the actuator, the input and output signals of the prime mover, the temperature gradient data of the high-temperature or medium-temperature energy storage chamber, and the real-time offset of the molten salt scaling factor in the speed control system device are collected;

[0079] A multi-objective rolling optimization algorithm is used to iteratively calculate the input and output signals of the regulator, the input and output signals of the actuator, the input and output signals of the prime mover, and the energy storage and scaling coupling parameters, so as to identify the following in the speed control system model: the regulator response hysteresis parameter, the actuator nonlinear characteristic parameter, the dynamic compensation coefficient of the phase change material heat capacity attenuation, and the coupling weight of the scaling factor and the heat capacity attenuation;

[0080] The objective function of the multi-objective rolling optimization algorithm is:

[0081] min(w1*E peak load regulation+w2*R power abandonment+w3*D thermal attenuation);

[0082] E peak regulation is the load tracking error, R curtailment is the wind and solar curtailment rate, and D thermal attenuation is the thermal capacity attenuation deviation; weights w1=0.4, w2=0.35, w3=0.25;

[0083] The corrected parameters are transmitted to the closed-loop control module.

[0084] The parameter identification module first obtains the basic characteristic parameters of the speed control system through static testing. During this static test, a constant load input signal is applied to the speed control system, and the actuator's displacement saturation point is recorded as the limiting parameter. When the input signal amplitude is less than the set threshold, the actuator's non-responsive interval is detected to determine the dead zone parameter. Simultaneously, the heat capacity loss rate of the phase change material in the energy storage chamber is measured under a constant temperature environment, and the initial heat capacity attenuation coefficient is calculated. These static parameters are then input into the speed control system model as basic constraints to establish the initial model framework.

[0085] Dynamic disturbance tests are conducted over a wide load range to collect multi-dimensional dynamic response data. The test selects multiple operating points within the range from rated load to the lower limit of deep peak regulation, and injects a step disturbance signal to trigger a sudden change in the main steam pressure. The command input signal and feedback output signal of the regulator are collected synchronously, and the control voltage and displacement output signal of the actuator are recorded to obtain the power input command and actual output response of the prime mover. Simultaneously, the temperature gradient distribution data of the high-temperature and medium-temperature energy storage chambers are monitored, and the molten salt scaling factor offset is obtained in real time through electrochemical sensors. All signal acquisition processes use a time domain alignment mechanism to ensure data timing consistency.

[0086] A multi-objective rolling optimization algorithm is used to iteratively calculate the collected data. The algorithm constructs a solution space that includes parameters such as the regulator's hysteresis time constant and the actuator's nonlinear friction coefficient. In each iteration, the regulator's actual input and output signals are compared with the model's predicted values. The hysteresis loop characteristics of the actuator's displacement response are analyzed, the phase deviation of the prime mover's power response is analyzed, and the coupling relationship between the energy storage chamber temperature gradient change rate and the scaling factor offset is combined. The objective function comprehensively evaluates three indicators: load tracking accuracy, renewable energy absorption efficiency, and thermal capacity attenuation compensation. The priorities of different optimization objectives are balanced through weight allocation.

[0087] The algorithm outputs four types of correction parameters: a regulator response lag parameter characterizing the time delay in command transmission; an actuator nonlinearity parameter reflecting the impact of friction on displacement accuracy; a dynamic compensation coefficient for the phase-change material's heat capacity decay quantifying the rate of heat loss; and a coupling weight between the scaling factor and heat capacity decay describing the impact of scaling deposition on heat transfer efficiency. These parameters are transmitted to the closed-loop control module via a data encapsulation protocol for real-time updating of the control strategy.

[0088] The parameter transmission process utilizes an industrial bus communication protocol, with a checksum implemented at the data transmission layer. Correction parameters are written to a shared memory area in a structured data format, and the closed-loop control module accesses this parameter data through address mapping. The transmission protocol includes data type identification, timestamp, and checksum fields to ensure the integrity and timeliness of parameter transmission.

[0089] Specifically, in the test system for correcting mismatch in the speed regulation model of a thermal power unit with a wide load range described in the present invention, the closed-loop control module generates a throttle valve opening instruction based on the corrected regulator response hysteresis parameter and actuator nonlinear characteristic parameter transmitted by the parameter identification module to adjust the turbine power output and suppress overregulation peaks caused by sudden load changes;

[0090] At the same time, based on the dynamic response parameters of the prime mover corrected by the parameter identification module, a combustion control signal is generated to adjust the fuel distribution ratio of the burner.

[0091] The closed-loop control module receives the corrected model parameters transmitted by the parameter identification module, including the regulator response lag parameter and the actuator nonlinear characteristic parameters. This module constructs the throttle valve opening feedforward compensation logic based on the regulator response lag parameters. Using a phase advance algorithm, it predicts the timing of throttle valve actuation during sudden load changes and generates a lead compensation command. The actuator nonlinear characteristic parameters are used in friction compensation calculations. A preset dead-band inverse model is used to correct displacement command deviations to offset the effects of mechanical transmission clearance on throttle valve positioning accuracy. The compensated throttle valve opening command drives the hydraulic motor or servo motor to adjust the turbine high-pressure throttle valve opening ratio in real time, suppressing power overshoot caused by sudden changes in main steam pressure.

[0092] This module simultaneously generates combustion control signals based on the prime mover's dynamic response parameters, which are modified by the parameter identification module. These parameters, including the volumetric time constant and the fuel-to-power conversion coefficient, are used to construct a fuel allocation ratio calculation model. This model dynamically calculates the optimal fuel supply coefficient for each burner based on the current load gradient and grid frequency fluctuations. The combustion control signals are transmitted to the boiler control system via the industrial bus, adjusting the fuel valve opening and air flow ratio to dynamically match the fuel calorific value release rate with the turbine power demand.

[0093] The throttle control command and the combustion control signal are synchronized and time-sequenced. During load increases, the fuel distribution ratio adjustment signal precedes the throttle opening command output. During load decreases, the throttle opening command takes precedence to prevent sudden increases in steam pressure. This coordination mechanism is implemented through a timestamp synchronization protocol, with command priority identifiers set within the control cycle. Actuator response status is fed back in real time via displacement sensors, forming a closed-loop verification mechanism between command output and execution feedback.

[0094] Burner fuel allocation is adjusted using a hierarchical optimization strategy. The primary control layer calculates the total fuel demand based on the prime mover's dynamic response parameters. The secondary control layer uses the accumulator temperature gradient data to allocate fuel supply weights to each burner. Burners in the high-temperature zone prioritize responding to load increases, while burners in the low-temperature zone take on load reduction adjustments. The allocation results are converted into a pulse-width modulated signal, which drives the fuel control valve actuator via a digital-to-analog converter.

[0095] This module's output signals form a closed-loop interaction with thermal system components. Changes in throttle valve opening are fed back to the main steam pressure monitoring node of the dynamic signal acquisition device, and burner status is fed back to the temperature monitoring node. This feedback data is used in parameter optimization calculations for the next control cycle, forming a closed-loop control chain from parameter modification to execution verification.

[0096] Specifically, the test system for correcting mismatch of speed control model of thermal power unit in wide load range described in the present invention, the simulation verification module receives the measured throttle oscillation frequency and main steam pressure attenuation rate generated by the dynamic signal processing module, and compares them with the dynamic response of the simulation model output after correction by the parameter identification module, adjusts the characteristic curve calibration parameters in the speed control system model through the error feedback mechanism, and transmits the corrected grid-related safety margin assessment result to the closed-loop control module for optimizing the generation logic of the throttle opening instruction and the combustion control signal.

[0097] The simulation verification module receives measured data from the dynamic signal processing module, including the frequency spectrum characteristics of the throttle valve oscillation and the time series of the main steam pressure decay rate. It also simultaneously obtains the corrected simulation model dynamic response data output by the parameter identification module, including the predicted waveform of the throttle valve opening and the simulated curve of steam pressure variation. After the two types of data are input into the verification interface, time domain alignment is performed, unifying the data sampling base using a timestamp matching algorithm.

[0098] Comparative verification analysis is performed based on the aligned data. Time-domain comparative analysis of the phase deviation and amplitude differences of the throttle oscillation waveform is performed, while frequency-domain comparative analysis of the energy distribution characteristics of the oscillation fundamental frequency is performed. The main steam pressure decay rate comparison utilizes slope change rate differential calculation to identify deviations between characteristic points of the measured decay curve and the simulated curve. The difference quantification process generates an error indicator matrix, including three core indicators: phase lag coefficient, amplitude deviation, and decay rate offset.

[0099] An error feedback mechanism drives iterative optimization of the speed control system model parameters. A parameter adjustment vector is generated based on the error indicator matrix and input into the high-pressure cylinder flow characteristic curve calibration function. Parameter optimization of the characteristic curve focuses on the nonlinear gain coefficient and the volume-time constant weight, calculating parameter corrections using the gradient descent method. The simulation model is rerun after each iteration, forming a closed-loop error convergence verification loop.

[0100] The grid-related safety margin assessment is based on the optimized model output. The assessment model integrates the grid frequency stability indicator system and calculates three key indicators: the power fluctuation threshold, the primary frequency regulation response time margin, and the frequency deviation safety range. The assessment process correlates the stability of the throttle control with the grid frequency fluctuation constraints, outputting a quantitative safety margin coefficient and risk level indicator.

[0101] The evaluation results are transmitted to the closed-loop control module via the data bus. The transmission protocol includes fields for a timestamp, margin factor, and parameter version number. After analyzing the evaluation results, the closed-loop control module adjusts the phase compensation weight of the throttle command based on the safety margin factor, dynamically switching the combustion control strategy based on the risk level. Control logic optimization utilizes a rules-based engine architecture, using a decision tree algorithm to match command generation patterns under different operating conditions.

[0102] Verification data is stored in a historical database for trend analysis. The database stores error indicator matrices and parameter correction records by operating point, supporting model generalization capability assessment. Regular data mining analysis is performed to identify parameter correction patterns across a wide load range, generating model adaptive calibration recommendations that are fed back to the parameter identification module.

[0103] Specifically, the test system for mismatch correction of speed regulation model of thermal power units in a wide load range according to the present invention comprises the following steps:

[0104] Injecting a step disturbance signal into the speed control system device within a wide load range of 100% rated load deep peak regulation lower limit, and collecting the throttle valve opening oscillation waveform output by the actuator when the main steam pressure suddenly changes;

[0105] The frequency characteristics of the throttle opening oscillation waveform are extracted through spectrum analysis, and the frequency characteristics are input into the speed control system model. The dynamic response parameters of the prime mover corrected by the parameter identification module are checked, and the corrected prime mover model parameters are generated and transmitted to the closed-loop control module.

[0106] The dynamic disturbance test is designed to cover a wide load range, from rated load to the lower limit of deep peak shaving. Before the test, load test points were selected based on the operating condition distribution diagram, encompassing three characteristic ranges: high load, sliding pressure, and deep peak shaving. A steady-state operating benchmark was established for each test point. A standardized step disturbance signal was injected into the speed control system via a signal generator. The disturbance amplitude was dynamically adjusted based on the load range, with small disturbances applied in the high load range and large disturbances in the deep peak shaving range.

[0107] The actuator's output response signal is collected in real time by a high-precision displacement sensor. Sudden changes in main steam pressure trigger throttle valve opening oscillations, and the displacement sensor records the throttle valve displacement curve at a millisecond sampling frequency. The actuator's oil pressure feedback signal and prime mover speed fluctuation data are simultaneously collected to form a time series dataset with multiple coupled physical quantities. The signal acquisition process utilizes time-scale synchronization technology, binding all sensor data to a unified time base.

[0108] The oscillation waveform spectrum is analyzed using a fast Fourier transform algorithm. The time-domain signal of the throttle opening is converted to the frequency domain, and three key features are extracted: the amplitude of the fundamental frequency component, the energy distribution of the main harmonic components, and the oscillation attenuation coefficient. The fundamental frequency component corresponds to the dominant oscillation mode of the system, the harmonic components reflect the nonlinear friction characteristics, and the attenuation coefficient represents the system damping level. The spectral characteristic parameters are encapsulated as feature vectors and input into the prime mover dynamic response submodule of the speed control system model.

[0109] After receiving the spectral eigenvector, the parameter verification module performs dynamic response parameter matching on the prime mover. The deviation between the measured fundamental frequency component and the model's predicted frequency is calculated to verify the calibration accuracy of the volume time constant. The harmonic component energy distribution is used to verify the validity of the nonlinear friction coefficient, while the attenuation coefficient is used to verify the system's damping compensation parameters. The verification process generates parameter corrections to dynamically update the prime mover model's transfer function coefficients and nonlinear link gains.

[0110] The corrected prime mover model parameters are transmitted to the closed-loop control module via an industrial communication protocol. Data transmission utilizes a publish-subscribe model, with the parameter identification module serving as the publisher and the closed-loop control module as the subscriber. The transmission protocol includes a parameter version identifier, an effective timestamp, and a checksum to prevent misalignment between control instructions and model parameter versions. The transmission period is dynamically adjusted based on the load change rate, shortening the transmission interval during periods of significant load fluctuations.

[0111] The closed-loop control module reconfigures the control algorithm after receiving the new parameters. The prime mover's volumetric time constant is used to optimize the feedforward compensation phase, and the friction coefficient weight is used to update the actuator deadband compensation strategy. A smooth transition mechanism is used during control strategy switching, and a parameter interpolation algorithm is used to avoid command jumps and maintain power output stability. The updated control results are fed back to the test system, forming a closed-loop verification chain.

[0112] Specifically, the test system for correcting mismatch of speed regulation model of thermal power unit in wide load range described in the present invention, the dynamic signal processing module performs spectrum correction processing on the throttle opening signal transmitted by the dynamic signal acquisition device, eliminates high-frequency noise interference introduced in the sensor acquisition process through digital filtering, generates a corrected throttle opening signal, and inputs the corrected throttle opening signal into the closed-loop control module, which is used to generate a throttle opening compensation instruction based on the regulator response lag parameter corrected by the parameter identification module, and adjust the turbine throttle opening.

[0113] The dynamic signal processing module receives the raw throttle valve opening signal transmitted by the dynamic signal acquisition device. This signal contains high-frequency noise components introduced by the sensor acquisition process, primarily in the form of random pulse interference and power frequency harmonic noise. The module uses digital filtering technology for preprocessing, designing a Butterworth low-pass filter to remove noise components above the system characteristic frequency. The filter cutoff frequency is set based on the natural frequency of the turbine's mechanical structure, retaining the low-frequency, effective signals related to the throttle valve's dynamic response.

[0114] The filtered signal undergoes phase correction. The group delay characteristics of the sensor signal transmission link are analyzed, and a phase compensation transfer function is constructed. A zero-phase filtering algorithm eliminates the phase offset of the signal acquisition system, synchronizing the time domain characteristics of the corrected throttle opening signal with the actual mechanical displacement. The correction process utilizes a sliding window processing mechanism, outputting real-time displacement data synchronized with the control system clock.

[0115] After generating the corrected throttle opening signal, key dynamic features are extracted. Feature extraction encompasses three dimensions: displacement gradient, oscillation period statistics, and steady-state offset. The displacement gradient calculates the rate of change of the opening per unit time, the oscillation period calculates the interval between consecutive waveform peaks, and the steady-state offset records the position deviation under steady-state conditions. Feature data is encapsulated as a structured array and timestamped.

[0116] The closed-loop control module receives the corrected throttle opening signal and characteristic data. This module uses the regulator response lag parameters corrected by the parameter identification module and constructs a feedforward compensation model based on the throttle displacement gradient characteristics. The model calculates the phase lead compensation amount and generates a throttle opening compensation command. This compensation command is added to the basic control command, and the final execution command is output through the proportional-integral-derivative control algorithm.

[0117] The actuator drives the throttle valve in position based on the throttle valve opening compensation command. A displacement sensor provides real-time feedback on the actual opening position, forming a closed-loop verification circuit. The feedback data is then compared with the compensation command to calculate the deviation. When the deviation exceeds a set threshold, adaptive adjustment of the compensation parameters is triggered. This adjustment optimizes the phase advance based on the statistical characteristics of the oscillation period and corrects the zero drift compensation value based on the steady-state offset.

[0118] The corrected throttle opening signal is used to evaluate control strategies. Dynamic feature data is stored in a historical database for analysis of control effectiveness under different load conditions. Feature cluster analysis is performed regularly to identify the mapping patterns between control parameters and throttle response characteristics. Parameter optimization recommendations are generated and fed back to the parameter identification module, forming a closed-loop iterative link between signal processing and control optimization.

[0119] Specifically, the test system for correcting mismatch in the speed regulation model of a thermal power unit with a wide load range according to the present invention further includes:

[0120] The speed regulation system device is connected to a preset two-stage energy storage structure, which includes: a low-pressure energy storage chamber, a medium-pressure energy storage chamber and a high-pressure energy storage chamber;

[0121] The high-pressure energy storage chamber and the medium-pressure energy storage chamber are filled with high-temperature phase change materials, and the melting point of the high-temperature phase change materials is ≥300℃;

[0122] Low-pressure energy storage chamber, filled with medium-temperature phase change material, 200℃≤medium-temperature phase change material melting point<300℃;

[0123] The speed control system device includes a hydraulic regulator, a PID controller and an actuator, which is used to dynamically adjust the opening of the turbine valve to control the power output.

[0124] The speed control system integrates a two-stage energy storage structure, which consists of a high-pressure energy storage chamber, a medium-pressure energy storage chamber, and a low-pressure energy storage chamber to form a hierarchical thermal energy buffer unit. The high-pressure and medium-pressure energy storage chambers are filled with high-temperature phase change materials, whose phase change temperature threshold is set to no less than 300 degrees Celsius. The material composition is an inorganic salt composite matrix and has chemical stability in high-temperature environments. The low-pressure energy storage chamber is filled with medium-temperature phase change materials, whose phase change temperature threshold ranges from 200 to 300 degrees Celsius. It uses organic-inorganic hybrid composite materials to optimize thermal cycle stability. Each energy storage chamber is connected in parallel with the turbine steam pipeline, and a fin-type heat exchange structure is installed inside the cavity to enhance the heat exchange efficiency between the phase change material and the steam.

[0125] The hydraulic regulator receives the power command signal and converts it into a hydraulic pressure control variable. This device integrates an electro-hydraulic servo valve group to dynamically adjust the hydraulic oil flow output based on load demand. A PID controller receives the speed deviation signal and load feedforward command and generates a control voltage signal using a proportional-integral-differential (PID) algorithm. This control algorithm includes nonlinear gain compensation to accommodate parameter variations over a wide load range. The actuator utilizes a high-precision hydraulic motor to convert the control voltage signal into a mechanical displacement output, driving the turbine throttle valve opening adjustment.

[0126] The thermal management mechanism of the two-stage energy storage structure operates based on the thermodynamic properties of phase change materials. When the main steam pressure rises suddenly, the phase change material in the high-temperature energy storage chamber absorbs excess heat energy, slowing the rate of pressure rise. When the pressure drops suddenly, the medium-temperature energy storage chamber releases stored heat energy to compensate for the pressure drop. The heat transfer process is monitored in real time by a temperature gradient sensor, which collects the temperature difference between the high- and medium-temperature energy storage chambers as an input parameter for the heat capacity decay calculation. A thermal insulation layer is installed on the outer wall of the energy storage chamber to reduce environmental heat loss and maintain the temperature stability of the phase change material.

[0127] The speed control system's control chain forms a closed-loop regulation circuit. The hydraulic regulator's output is connected to the actuator's hydraulic drive unit, and the actuator's displacement feedback signal is fed back to the PID controller's input. The displacement feedback signal is collected by a linear variable differential transformer and converted into a standardized voltage signal by a signal conditioning circuit. The control loop includes a built-in limiter protection module, which constrains output commands based on displacement limits obtained from static testing to prevent mechanical overload. The deadband compensation module dynamically adjusts control gains based on nonlinear characteristic parameters to eliminate regulation lag caused by transmission backlash.

[0128] This structure optimizes the dynamic characteristics of power output through thermal and mechanical coupling. The thermal buffering effect of the phase change material smoothes steam pressure fluctuations and reduces the frequency of valve opening adjustment; the precise displacement control of the mechanical actuator synchronously responds to load changes. The temperature gradient data of the two-stage energy storage chamber is input into the dynamic signal processing module, which participates in the calculation of the heat capacity attenuation slope to form a thermal energy management closed loop. The state parameters of the actuator are fed back to the parameter identification module to continuously optimize the control algorithm parameters and maintain the speed regulation stability within the full load range.

[0129] Specifically, the test system for correcting mismatch in the speed regulation model of a thermal power unit with a wide load range according to the present invention further includes:

[0130] The multi-objective rolling optimization algorithm simultaneously optimizes peak shaving depth, equipment life, and wind and solar curtailment rates;

[0131] Dynamic binding of renewable energy consumption priorities to energy storage and release strategies;

[0132] and transmitting the corrected model parameters to the closed-loop control module;

[0133] The closed-loop control module generates a throttle opening instruction and a combustion control signal according to the corrected model parameters, drives the actuator to adjust the turbine power output, suppresses the over-modulation peak caused by the sudden change in load, and optimizes the burner fuel distribution to reduce the delay of the primary frequency modulation action.

[0134] A multi-objective rolling optimization algorithm constructs a three-dimensional solution space encompassing peak-shaving depth, equipment lifespan, and wind and solar curtailment rates. The peak-shaving depth metric is quantified using a load tracking capability assessment model, calculating the integral deviation between actual and target loads. The equipment lifespan metric is based on a material fatigue accumulation model, analyzing the number of stress cycles experienced by key components under varying operating conditions. The wind and solar curtailment rate metric is linked to grid dispatch instructions and measures the proportion of periods with limited renewable energy output. The algorithm utilizes a rolling time-domain optimization framework, dynamically adjusting objective weight coefficients within each optimization cycle to generate parameter correction vectors.

[0135] A dynamic binding mechanism is established between renewable energy consumption priorities and energy storage release strategies. When the power grid issues a high-proportion renewable energy consumption order, an optimization algorithm increases the weight coefficient of the wind and solar curtailment rate indicator, triggering an adjustment to the energy storage chamber release strategy. This strategy adjusts the operating temperature range of the phase change material to alter the distribution ratio of the thermal storage capacity of the energy storage chamber. High-pressure energy storage chambers prioritize energy release requirements, while medium-pressure storage chambers assume energy storage balancing tasks, forming a thermal energy dispatch logic that matches consumption priorities.

[0136] The corrected model parameters are transmitted to the closed-loop control module via the industrial bus. Data transmission utilizes a publish / subscribe model, with the parameter identification module acting as a data producer and publishing parameter update messages, while the closed-loop control module acts as a consumer and subscribes to these updates. The message structure includes a parameter version identifier, an effective timestamp, and a data checksum field to prevent inconsistencies between control instructions and model parameter versions. The transmission period is adaptively adjusted based on the load change rate, shortening the transmission interval during periods of significant load fluctuations.

[0137] The closed-loop control module analyzes the model parameters and reconstructs the control strategy. The regulator response lag parameters are used to construct a feedforward compensation model for the throttle valve opening, generating compensation commands using a phase advance algorithm. The actuator's nonlinear characteristic parameters are input into the friction compensation calculation unit, which uses a preset deadband inverse model to correct displacement deviations. The compensated throttle valve opening command drives the servo actuator, which adjusts the turbine's high-pressure throttle valve opening via a mechanical transmission device, suppressing power overshoot peaks caused by sudden changes in main steam pressure.

[0138] Combustion control signals are generated based on the dynamic response parameters of the prime mover. The volumetric time constant output by the parameter identification module is used to calculate the fuel supply gradient. This is then combined with the load change rate and steam pressure decay rate to dynamically allocate fuel to each burner. Signal output utilizes a hierarchical control architecture: the primary control layer generates the total fuel demand command, while the secondary control layer optimizes fuel allocation based on the accumulator temperature distribution. The control command is converted into a valve position signal and transmitted via the fieldbus to the combustion actuator for fuel adjustment, shortening the delay time of the primary frequency modulation operation.

[0139] This module establishes a control effectiveness feedback mechanism. The throttle position sensor collects actual valve opening position in real time, and the burner status monitoring unit provides fuel distribution data. This feedback signal is input into the dynamic signal acquisition device and used to optimize parameters for the next control cycle. The control strategy version and model parameter version are verified through timestamp matching, forming a closed-loop control chain from parameter update to execution verification.

[0140] Specifically, the test system for correcting mismatch in the speed regulation model of a thermal power unit with a wide load range according to the present invention further includes:

[0141] The simulation verification module compares the measured data generated by the dynamic signal processing module with the simulation model output corrected by the parameter identification module, dynamically adjusts the model characteristic curve calibration parameters through the error feedback mechanism, and verifies the real-time compensation model for the scaling factor of the thermoelectric decoupling heat exchanger:

[0142] δt=δ0×e-kt+β×ΔTmolten salt;

[0143] Where δ0 is the initial scaling factor, k is the attenuation coefficient, β is the temperature compensation coefficient, and ΔTmoltensalt is the molten salt temperature difference.

[0144] The simulation verification module receives the measured data set transmitted by the dynamic signal processing module, including the time-domain signal of the throttle valve opening oscillation waveform and the main steam pressure decay rate curve. It simultaneously obtains the simulation model's dynamic response data output by the parameter identification module, including the throttle valve opening prediction curve and the simulated steam pressure change rate. The two types of data are input into the comparative analysis interface, and time-domain synchronization and alignment are performed. A feature point matching algorithm is used to unify the time base and eliminate phase deviation caused by sampling frequency differences.

[0145] The comparative analysis process utilizes a multi-dimensional error quantification method. The comparison of throttle oscillation characteristics focuses on three core indicators: fundamental frequency component amplitude deviation, harmonic distortion, and decay time constant. The comparison of main steam pressure variation characteristics utilizes a dynamic time warping algorithm to calculate the morphological similarity between the measured and simulated curves. The error quantification results generate a feature difference matrix, which includes three evaluation parameters: amplitude deviation coefficient, phase lag, and morphological distortion.

[0146] An error feedback mechanism drives iterative optimization of model parameters. The characteristic difference matrix is ​​input into the high-pressure cylinder flow characteristic curve parameter adjustment module, which includes a nonlinear gain coefficient calibration function and a volume-time constant weighting algorithm. Parameter adjustment utilizes an adaptive gradient descent method, correcting the gain curve slope based on the amplitude deviation coefficient and optimizing the time constant weighting ratio based on the phase lag. The simulation model is rerun after each parameter update, forming a closed-loop error convergence verification loop.

[0147] The scaling factor compensation model for thermoelectric decoupled heat exchangers is independently validated. This validation process builds a scaling factor trend analysis model based on molten salt temperature differential monitoring data and historical scaling deposition records. The validation module compares the theoretical scaling factor calculated in real time with the measured offset and evaluates the effectiveness of the compensation model through residual analysis. When the residual exceeds the allowable threshold, the adaptive compensation factor adjustment mechanism is triggered, updating the attenuation coefficient and temperature compensation coefficient.

[0148] The grid-related safety margin assessment is based on the output of the optimized model. The assessment model integrates into the grid frequency stability analysis framework and calculates three metrics: the permissible power fluctuation range, the frequency regulation response time window, and the frequency excursion safety margin. The assessment process correlates the stability of the throttle control with grid frequency constraints, outputting a quantitative safety level coefficient and a risk identification code. The safety level coefficient reflects the unit's grid adaptability over a wide load range, while the risk identification code identifies the type of operational risk under critical operating conditions.

[0149] The assessment results are transmitted to the closed-loop control module via an industrial communication protocol. Data transmission utilizes a timestamped message structure, including fields for the safety level coefficient, risk identification code, and effective operating condition range. After parsing the message, the closed-loop control module adjusts the throttle feedforward compensation strength based on the safety level coefficient and switches the combustion control strategy based on the risk identification code. Control logic reconstruction utilizes a rules engine architecture, using a decision tree model to match command generation patterns for different safety levels.

[0150] Historical validation data is stored in a knowledge base system. This knowledge base stores error quantification records, parameter adjustment trajectories, and safety assessment results by load interval, supporting data trend mining and analysis. Cluster analysis algorithms are regularly executed to identify the mapping relationship between parameter modification patterns and operating condition characteristics. This generates a model generalization capability assessment report that is fed back to the parameter identification module, forming a closed-loop iterative system for verification and optimization.

[0151] The technical features involved in the technical solution of the present invention are explained as follows:

[0152] Correlation analysis logic of dynamic signal processing module:

[0153] This module implements multi-parameter coupling analysis through a hierarchical signal feature extraction method:

[0154] Main steam pressure sudden change identification: adopt the pressure change rate threshold judgment mechanism, combined with the time window integration algorithm to filter out instantaneous interference and accurately capture effective pressure mutation events.

[0155] Quantification of oscillation characteristics: Perform spectral energy analysis on the filtered displacement signal to extract the fundamental frequency amplitude and attenuation coefficient, eliminating the interference effect of mechanical transmission noise.

[0156] Heat capacity attenuation correlation modeling: Generate dynamic attenuation parameters based on the proportional relationship between the temperature difference in the energy storage chamber and the heat capacity loss, reflecting the changing trend of the phase change material's energy storage / release efficiency.

[0157] Multi-objective optimization mechanism of parameter identification module:

[0158] This module uses a constrained hierarchical solution strategy to achieve model mismatch correction:

[0159] Objective function design: Load tracking error, wind and solar power curtailment rate, and thermal capacity attenuation deviation are combined into a comprehensive optimization target according to preset weights. The weight distribution reflects the balance between the grid frequency regulation demand and equipment life.

[0160] Rolling time domain optimization: Parameters are solved in segments in dynamic disturbance tests, and the optimization results of the previous segment serve as constraints for the subsequent segment to adapt to the time-varying characteristics of parameters under varying working conditions.

[0161] Physical parameter output: The identification results directly point to measurable physical quantities (such as regulator lag time and actuator friction coefficient), avoiding abstract mathematical descriptions.

[0162] Physical compensation logic of closed-loop control module:

[0163] Feedforward compensation of throttle opening: Convert the regulator lag time into phase advance, generate compensation instructions through transfer function reconstruction, and convert them into differential equations executable by PID controller during specific implementation.

[0164] Dynamic adjustment of fuel distribution: The fuel supply gradient is set based on the prime mover volume time constant, and the rate of change of power demand is matched through the burner valve opening ratio control.

[0165] Engineering verification path for simulation verification modules:

[0166] Dynamic response comparison method: The waveform morphology similarity algorithm is used to align the measured and simulated curves, and the residual calculation results drive the iteration of the high-pressure cylinder flow characteristic parameters.

[0167] Implementation of scaling compensation model: The initial scaling factor is obtained through offline calibration, the attenuation coefficient and temperature compensation coefficient are regressed and fitted based on long-term operating data, and the model output is an executable compensation instruction.

[0168] Network security constraint conversion: Convert indicators such as power fluctuation threshold and frequency modulation time margin into rate limit parameters for control instructions.

[0169] Description of the feasibility of the technical solution:

[0170] Parameter acquisition path: The regulator hysteresis time is measured through a step response test; the molten salt scaling factor is monitored online using an industrial-grade electrochemical impedance spectroscopy instrument;

[0171] Algorithm engineering implementation: Signal processing modules can be deployed on DSP chips; optimization algorithms can be run on industrial computers; and control logic can be embedded in the unit DCS system.

[0172] Technical effect verification: load tracking error, primary frequency modulation delay and other indicators are tested according to DL / T standards, and the results are reproducible.

[0173] The embodiments of the present invention are as follows:

[0174] The test system of the present invention is applied to the deep peak regulation scenario of a 600MW coal-fired unit, and the speed regulation model mismatch correction is achieved through the following steps:

[0175] System construction: The thermal system components simulate the thermal cycle of boilers and steam turbines. The main steam pressure sensor (range 0-25MPa, accuracy ±0.1%) and temperature sensor (range 0-600℃, accuracy ±1℃) collect steam parameters in real time with a sampling frequency of 1kHz.

[0176] The speed control system is equipped with an electro-hydraulic actuator (stroke 0-300mm, resolution 0.01mm) to drive the turbine valve, and the speed measurement module has an accuracy of ±0.05% of the rated speed.

[0177] The two-stage energy storage structure includes: high-pressure / medium-pressure energy storage chambers are filled with high-temperature nitrate phase change material (heat capacity ≥ 1.5kJ / kg·K) with a melting point of 320°C;

[0178] The low-pressure energy storage chamber is filled with a medium-temperature fatty acid-based phase change material (heat capacity ≥ 2.0 kJ / kg·K) with a melting point of 250°C.

[0179] The energy storage chamber temperature difference sensor (range 0-200℃, accuracy ±0.5℃) monitors the temperature gradient.

[0180] Dynamic signal processing:

[0181] The dynamic signal acquisition device collects the throttle opening oscillation signal (bandwidth 0-50Hz) and eliminates high-frequency noise through a Butterworth low-pass filter (cut-off frequency 15Hz).

[0182] Identify main steam pressure sudden change events (>1.2MPa / s), and correlate them with the throttle valve opening fluctuation amplitude (>6%) and the high-temperature energy storage chamber heat capacity attenuation slope (calculation period 100ms).

[0183] The molten salt scaling factor is monitored in real time by an impedance analyzer (frequency range 10 Hz-100 kHz) and fused with the heat capacity decay data to generate a dynamic response database.

[0184] Parameter identification and correction:

[0185] The static test calibrates the actuator dead zone (1.2mm) and limit range (0-100% opening).

[0186] A step disturbance (amplitude 8% of rated load) is injected in the load range of 40%-100%, and the regulator response lag data (typical value 120ms) is collected.

[0187] The multi-objective optimization algorithm is iteratively calculated with weights (w1=0.4,w2=0.35,w3=0.25) and outputs:

[0188] Actuator nonlinear friction coefficient compensation (0.15-0.35);

[0189] Thermal capacity attenuation dynamic compensation coefficient (1.05-1.25);

[0190] Scaling and heat capacity coupling weight (0.45).

[0191] Closed-loop control execution:

[0192] Feed-forward compensation of the throttle opening: Based on the regulator lag time (95ms after correction), a compensation command with a phase advance of 25° is generated, which reduces the power overshoot peak by 45% when the load changes suddenly (the measured overshoot is reduced from 8.5% to 4.7%).

[0193] Combustion control optimization: Based on the prime mover volume time constant (12s after correction), the fuel distribution ratio was adjusted (the fuel proportion of the high-temperature zone burner was increased to 75%), and the delay of the first frequency modulation action was shortened to 1.8s (original value 4.5s).

[0194] Simulation verification and security enhancement:

[0195] The measured main steam pressure decay rate (0.85 MPa / s) was compared with the model output (0.82 MPa / s), and the gain coefficient of the high-pressure cylinder flow characteristic curve was optimized (adjustment amount +8%) through the residual sum of squares.

[0196] Verification of the scaling compensation model: initial scaling factor δ0 = 1.25, attenuation coefficient k = 0.003s⁻¹, temperature compensation coefficient β = 0.02, and compensation error ≤ 5% when the molten salt temperature difference ΔT = 35°C.

[0197] Grid-related safety margin assessment: power fluctuation threshold ±12MW (±2% rated value), frequency regulation response time margin 1.6s. The control module limits the throttle opening change rate to ≤3% / s accordingly.

[0198] Designed for wide-load peak-shaving scenarios in thermal power plants, the test system of this invention integrates thermal system components, a speed control system, a dynamic signal acquisition device, and a control device. This system addresses the existing problem of speed control system performance test deviations caused by dynamic response mismatch during deep peak-shaving of thermal power plant steam turbines. These deviations manifest as inaccurate controller response characteristics, reduced speed regulation accuracy, and dynamic fluctuations in combustion parameters. The control device establishes real-time communication links with various components, forming a closed-loop testing and correction architecture.

[0199] The thermal system components simulate the actual thermal cycle of a thermal power plant. The main steam parameters generated by the boiler are collected in real time by high-precision sensors at a sampling frequency of 1 kHz, covering the entire peak-shaving range from rated load to the lower limit of deep peak-shaving (e.g., 30% of rated load). The speed control system integrates a hydraulic regulator, a PID controller, and an actuator. The actuator drives the turbine throttle valve opening to control power output. A displacement sensor collects the throttle valve opening signal (with a resolution of 0.01 mm) and simultaneously records the speed deviation signal (with an accuracy of ±0.1 rpm). The speed control system is connected to a pre-designed two-stage energy storage structure: the high-pressure and medium-pressure energy storage chambers are filled with high-temperature phase change materials (such as nitrate complexes) with a melting point of 300°C or higher, while the low-pressure energy storage chamber is filled with medium-temperature phase change materials (such as organic-inorganic hybrids) with a melting point of 200-300°C. The temperature gradient in the energy storage chambers is measured using a multi-point temperature sensor array (with a temperature differential accuracy of ±0.5°C). The molten salt scaling factor is indirectly measured using electrochemical impedance spectroscopy with a sampling period of 100 ms.

[0200] The dynamic signal processing module preprocesses the sensor's raw signal: a Butterworth low-pass filter with a cutoff frequency of 50 Hz is used to eliminate high-frequency noise and retain the fundamental frequency component of the throttle oscillation (typical range 2-15 Hz). The module analyzes the main steam pressure change rate (pressure change per unit time dP / dt), the throttle opening fluctuation amplitude (maximum displacement deviation within the oscillation period), and the energy storage chamber heat capacity attenuation slope (calculated based on the ratio of the temperature difference ΔT to the heat capacity loss ΔQ). The module also integrates the molten salt scaling factor change rate to identify the dynamic correlation between the main steam pressure sudden change (threshold >1 MPa / s), the throttle opening oscillation (amplitude >5%), and the heat capacity attenuation, and generates a time series dynamic response database.

[0201] The parameter identification module uses static testing to determine the actuator displacement deadband (0.5-2mm), limiting range (0-100% opening), and initial thermal capacity attenuation coefficient of the phase change material. Dynamic disturbance testing injects a step disturbance (amplitude 5-10% of rated load) across a wide load range, collecting regulator input / output signals, actuator displacement response, and energy storage chamber temperature gradient. A multi-objective rolling optimization algorithm is used, iteratively calculating the objective function min(0.4 load tracking error + 0.35 wind and solar curtailment rate + 0.25 * thermal capacity attenuation deviation). The module identifies the regulator response lag time (50-200ms), actuator nonlinear friction coefficient, dynamic compensation coefficient for thermal capacity attenuation of the phase change material, and the coupling weight (0.2-0.8) between scaling and thermal capacity attenuation. The module then outputs the corrected speed regulation model parameters.

[0202] The closed-loop control module generates control instructions based on the correction parameters: a phase advance algorithm (advance angle 10°-30°) is used according to the regulator lag time to generate a throttle opening feedforward compensation instruction to suppress the power overmodulation peak caused by sudden load changes (overmodulation is reduced by more than 40%); the fuel distribution ratio is dynamically adjusted according to the dynamic response parameters of the prime mover (such as a high / low temperature burner weight ratio of 3:1), and the air supply ratio signal is combined to shorten the delay of a single frequency modulation action to within 2 seconds; the throttle displacement feedback and burner status feedback are used to optimize the instructions for the next cycle to form a closed-loop control.

[0203] The simulation verification module compares the measured throttle oscillation frequency and main steam pressure decay rate with the corrected model output. It dynamically adjusts the high-pressure cylinder flow characteristic curve parameters through an error feedback mechanism (minimizing the sum of squared residuals). The effectiveness of the fouling compensation model for the thermoelectric decoupled heat exchanger is verified, outputting real-time compensation based on the initial fouling factor δ0, the decay coefficient k, and the molten salt temperature difference ΔT. A grid-related safety margin assessment calculates the power fluctuation threshold (±2% of rated power), the frequency modulation response time margin (>1.5s), and the frequency deviation safety range (49.8-50.2Hz). These assessment results are fed back to the closed-loop control module to optimize the command logic.

[0204] Through dynamic signal correlation analysis, multi-objective parameter identification and closed-loop verification mechanism, this system can achieve real-time correction of speed regulation model mismatch under wide load conditions, reduce the load tracking error in the entire peak-shaving interval to within 1%, and reduce the furnace pressure fluctuation amplitude by more than 50%, thus meeting the requirements of deep peak-shaving for dynamic response accuracy and grid security.

[0205] To address the problem of poor coordinated control quality, the present invention uses a dynamic signal processing module to analyze the correlation characteristics between sudden changes in main steam pressure and oscillations in the throttle valve opening in real time, and constructs a dynamic response database including time series. Based on this database, the parameter identification module uses a multi-objective rolling optimization algorithm to identify the regulator response lag parameters and the nonlinear characteristic parameters of the actuator in the speed regulation system, and correct the model mismatch under wide load conditions. The closed-loop control module generates a throttle valve opening feedforward compensation instruction based on the corrected parameters, offsets the control lag effect through the phase advance compensation mechanism, suppresses the power overregulation peak caused by sudden load changes, and achieves accurate load tracking in the entire peak regulation range.

[0206] To address the degraded primary frequency regulation performance, the system utilizes the thermal buffering properties of a two-stage energy storage structure to smooth steam pressure fluctuations. High-temperature phase change materials are used in the high- and medium-pressure energy storage chambers, while medium-temperature phase change materials are used in the low-pressure energy storage chamber. These materials absorb and release thermal energy through phase change. The parameter identification module, combined with energy storage chamber temperature gradient data, corrects the prime mover's dynamic response parameters and heat capacity attenuation coefficient. Based on these corrected parameters, the closed-loop control module generates combustion control signals and dynamically optimizes the fuel distribution ratio and air volume ratio to align the combustion response rate with grid frequency fluctuations, thereby reducing the delay in primary frequency regulation.

[0207] To address combustion stability issues, the simulation verification module compares measured data with model outputs and dynamically adjusts the calibration parameters of the high-pressure cylinder flow characteristic curve through an error feedback mechanism. The fouling factor compensation model for the thermoelectric decoupled heat exchanger is also verified. This model integrates the initial fouling factor, attenuation coefficient, and molten salt temperature difference compensation to correct heat transfer efficiency parameters in real time. The optimized grid-related safety margin assessment results are fed back to the closed-loop control module, which reconfigures the throttle opening command and fuel distribution strategy to maintain furnace pressure stability under different load conditions and eliminate combustion fluctuations caused by sudden changes in main steam parameters.

Claims

1. A test system for speed regulation model mismatch correction of thermal power units in a wide load range, characterized by: include: Thermal system components, speed control system devices, dynamic signal acquisition devices and control devices, the control device establishes communication connections with the thermal system components, speed control system devices and dynamic signal acquisition devices; The dynamic signal acquisition device includes a high-precision sensor that collects main steam pressure, temperature, valve opening, speed deviation signal and energy storage chamber temperature gradient data in real time and transmits it to the control device; The control device includes a dynamic signal processing module, a parameter identification module, a closed-loop control module and a simulation verification module; The dynamic signal processing module receives the sensor signal, analyzes the main steam parameter change trend, the valve oscillation characteristics and the phase change material heat capacity attenuation characteristics, generates a dynamic response database including a time series, and transmits the dynamic response database to the parameter identification module; The parameter identification module identifies the regulator response hysteresis parameters, actuator nonlinear characteristic parameters, prime mover dynamic response parameters, and phase change material heat capacity attenuation coefficient in the speed control system device through a multi-objective rolling optimization algorithm based on the dynamic disturbance test data in the dynamic response database, corrects the model mismatch of the preset speed control system model under a wide load condition, and obtains the corrected parameters to generate a throttle opening instruction and a combustion control signal; The closed-loop control module generates a throttle opening instruction and a combustion control signal based on the corrected parameters, and drives the speed control system device to perform turbine power adjustment and burner fuel distribution. The simulation verification module receives real-time data from the dynamic signal acquisition device and the corrected model output of the parameter identification module, compares the measured response with the simulation result through the error feedback mechanism, dynamically adjusts the calibration parameters of the high-pressure cylinder flow characteristic curve, calculates the power fluctuation allowable threshold, the frequency modulation response time margin and the frequency deviation safety range based on the comparison error data and the fouling compensation model verification result, generates a grid-related safety margin assessment result, and feeds the assessment result back to the closed-loop control module to optimize the constraint conditions of the throttle opening instruction and the combustion control signal.

2. The test system for speed regulation model mismatch correction of thermal power units in a wide load range according to claim 1 is characterized in that: The dynamic signal processing module analyzes the nonlinear fluctuation characteristics in the sensor signal transmitted by the dynamic signal acquisition device, synchronously integrates the energy storage chamber temperature gradient data and the molten salt scaling factor change rate, identifies the dynamic correlation between the main steam pressure sudden change, the throttle opening oscillation and the phase change material heat capacity attenuation, and generates the dynamic response database including the time series of the main steam pressure change rate, the throttle opening fluctuation amplitude, the speed deviation, the high-temperature phase change material heat capacity attenuation slope and the molten salt scaling factor dynamic offset; The heat capacity attenuation slope of the high-temperature phase change material is calculated by the following formula: α=-(ΔT gradient / ΔtΔQ heat capacity); ΔQ heat capacity is the heat capacity loss per unit time of the energy storage cavity, ΔT gradient is the temperature difference between the high / medium temperature energy storage cavity, and Δt is the sampling period; The correlation data in the dynamic response database is input into the parameter identification module to correct the regulator response hysteresis parameters and the phase change material thermodynamic performance and scaling coupling characteristic parameters in the preset speed control system model.

3. The test system for speed regulation model mismatch correction of thermal power units in a wide load range according to claim 2 is characterized in that: The parameter identification module is configured to: Obtain the limiting parameters, dead zone parameters and initial heat capacity attenuation coefficient of the phase change material of the energy storage cavity in the speed regulation system device through static tests, and input them into the speed regulation system model; Based on the dynamic disturbance test, the input and output signals of the regulator, the input and output signals of the actuator, the input and output signals of the prime mover, the temperature gradient data of the high-temperature or medium-temperature energy storage chamber, and the real-time offset of the molten salt scaling factor in the speed control system device are collected; A multi-objective rolling optimization algorithm is used to iteratively calculate the input and output signals of the regulator, the input and output signals of the actuator, the input and output signals of the prime mover, and the energy storage and scaling coupling parameters, so as to identify the following in the speed control system model: the regulator response hysteresis parameter, the actuator nonlinear characteristic parameter, the dynamic compensation coefficient of the phase change material heat capacity attenuation, and the coupling weight of the scaling factor and the heat capacity attenuation; The objective function of the multi-objective rolling optimization algorithm is: min(w1*E peak load regulation+w2*R power abandonment+w3*D thermal attenuation); E peak regulation is the load tracking error, R curtailment is the wind and solar curtailment rate, and D thermal attenuation is the thermal capacity attenuation deviation; weights w1=0.4, w2=0.35, w3=0.25; The corrected parameters are transmitted to the closed-loop control module.

4. The test system for speed regulation model mismatch correction of thermal power units in a wide load range according to claim 3 is characterized in that: The closed-loop control module generates a throttle valve opening instruction based on the corrected regulator response hysteresis parameters and actuator nonlinear characteristic parameters transmitted by the parameter identification module to adjust the turbine power output and suppress the over-regulation peak caused by sudden load changes; At the same time, based on the dynamic response parameters of the prime mover corrected by the parameter identification module, a combustion control signal is generated to adjust the fuel distribution ratio of the burner.

5. The test system for mismatch correction of speed regulation model of thermal power units in a wide load range according to claim 4 is characterized in that: The simulation verification module receives the measured throttle oscillation frequency and main steam pressure attenuation rate generated by the dynamic signal processing module, compares them with the dynamic response output by the simulation model corrected by the parameter identification module, adjusts the characteristic curve calibration parameters in the speed control system model through the error feedback mechanism, and transmits the corrected grid-related safety margin assessment results to the closed-loop control module for optimizing the generation logic of the throttle opening command and the combustion control signal.

6. The test system for mismatch correction of speed regulation model of thermal power units in a wide load range according to claim 5, characterized in that: The dynamic disturbance test includes: Injecting a step disturbance signal into the speed control system device within a wide load range of 100% rated load deep peak regulation lower limit, and collecting the throttle valve opening oscillation waveform output by the actuator when the main steam pressure suddenly changes; The frequency characteristics of the throttle opening oscillation waveform are extracted through spectrum analysis, and the frequency characteristics are input into the speed control system model. The dynamic response parameters of the prime mover corrected by the parameter identification module are checked, and the corrected prime mover model parameters are generated and transmitted to the closed-loop control module.

7. The test system for speed regulation model mismatch correction of thermal power units in a wide load range according to claim 6, characterized in that: The dynamic signal processing module performs spectrum correction processing on the throttle opening signal transmitted by the dynamic signal acquisition device, eliminates high-frequency noise interference introduced in the sensor acquisition process through digital filtering, generates a corrected throttle opening signal, and inputs the corrected throttle opening signal into the closed-loop control module for generating a throttle opening compensation instruction based on the regulator response lag parameter corrected by the parameter identification module to adjust the turbine throttle opening.

8. The test system for mismatch correction of speed regulation model of thermal power units in a wide load range according to claim 7, characterized in that: Also includes: The speed regulation system device is connected to a preset two-stage energy storage structure, which includes: a low-pressure energy storage chamber, a medium-pressure energy storage chamber and a high-pressure energy storage chamber; The high-pressure energy storage chamber and the medium-pressure energy storage chamber are filled with high-temperature phase change materials, and the melting point of the high-temperature phase change materials is ≥300℃; Low-pressure energy storage chamber, filled with medium-temperature phase change material, 200℃≤medium-temperature phase change material melting point<300℃; The speed control system device includes a hydraulic regulator, a PID controller and an actuator, which is used to dynamically adjust the opening of the turbine valve to control the power output.

9. The test system for mismatch correction of speed regulation model of thermal power units in a wide load range according to claim 8, characterized in that: Also includes: The multi-objective rolling optimization algorithm simultaneously optimizes peak shaving depth, equipment life, and wind and solar curtailment rates; Dynamic binding of renewable energy consumption priorities to energy storage and release strategies; and transmitting the corrected model parameters to the closed-loop control module; The closed-loop control module generates a throttle opening instruction and a combustion control signal based on the corrected model parameters, drives the actuator to adjust the turbine power output, suppresses the over-modulation peak caused by sudden load changes, and optimizes the burner fuel distribution to reduce the delay of the primary frequency modulation action.

10. The test system for mismatch correction of speed regulation model of thermal power units in a wide load range according to claim 9, characterized in that: Also includes: The simulation verification module compares the measured data generated by the dynamic signal processing module with the simulation model output corrected by the parameter identification module, dynamically adjusts the model characteristic curve calibration parameters through the error feedback mechanism, and verifies the real-time compensation model for the scaling factor of the thermoelectric decoupling heat exchanger: δt=δ0× +β×ΔT molten salt; Where δ0 is the initial scaling factor, k is the attenuation coefficient, β is the temperature compensation coefficient, and ΔTmoltensalt is the molten salt temperature difference.

Images