An optimization method for parameters of a steam turbine power-frequency electro-hydraulic control system

Abstract

The present invention provides an optimization method for the parameters of a steam turbine power-frequency electro-hydraulic control system. The present invention relates to the technical field of system parameter optimization. The present invention trains an adjustment prediction model through historical speed curves and power adjustment amounts, obtains a predicted power adjustment amount according to the real-time speed curve, obtains a flow opening formula and steam energy according to the relationship between the opening of the valve and the flow rate, analyzes the formula of the steam energy in MATLAB to obtain the corresponding relationship between the opening interval and the selection interval, constructs an opening score to obtain the opening interval with the most accurate control of the steam energy by the valve opening, constructs a change index according to the deviation between the actual power adjustment amount and the predicted power adjustment amount, and judges the pressure required for secondary frequency modulation and the opening interval of the valve according to the relationship between the change index and the index threshold. The present invention realizes the precise control of the steam turbine rotor, thereby improving the control performance of the power-frequency electro-hydraulic system.

Description

Technical Field

[0001] The present invention relates to the technical field of system parameter optimization, in particular to a method for optimizing parameters of a steam turbine power frequency electro-hydraulic control system. Background Art

[0002] With the continuous increase in electricity demand and the increase in the proportion of renewable energy, frequency deviations in the power grid often occur. When the power grid has a frequency deviation, the generator set will actively perform a frequency adjustment, and then the AGC system will issue instructions to each unit for a secondary frequency adjustment based on the specific deviation situation, requiring the unit's power generation status to make corresponding changes, thereby adjusting the frequency deviation in the power grid. The steam turbine is an important power generation equipment in the thermal power unit. The power generation power is directly affected by the rotor speed in the steam turbine unit, and the rotor speed is controlled by the steam flow to realize the conversion of steam energy and power generation power. In the process of secondary frequency regulation, the AGC system will directly take over the power frequency electro-hydraulic system to adjust the power generation power of the steam turbine, and then the power frequency electro-hydraulic system controls the pressure of the steam chamber and the valve to adjust the speed of the steam turbine unit. The control accuracy of the pressure and valve for the flow directly affects the performance of the power frequency electro-hydraulic system for the speed regulation ability, thereby affecting the process of secondary frequency regulation.

[0003] In the prior art, publication number CN103401256A discloses a method for optimizing the parameters of the power frequency electro-hydraulic control system of a large steam turbine in a thermal power plant, the impact of the power frequency electro-hydraulic control system of a steam turbine generator set on the stability of the power system, and the power frequency electro-hydraulic control system parameters are obtained through an intelligent optimization algorithm in combination with the primary frequency regulation performance of the unit and the dynamic stability margin requirements of the system.

[0004] According to public documents, the optimization of PID control parameters in the primary frequency modulation process of the steam turbine unit was achieved. The PID control parameters belong to the parameter optimization of the power frequency electro-hydraulic system from the data level, but the secondary frequency modulation process was not optimized. In addition, the two key physical control parameters of the relevant pressure parameters and valve opening in the frequency modulation process were not considered. The problem of inaccurate control caused by the hardware's own characteristics during the frequency modulation process could not be avoided, and the parameters of the power frequency electro-hydraulic system were not optimized from the physical level.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the Invention

[0006] The object of the present invention is to provide a method for optimizing parameters of a steam turbine power frequency electro-hydraulic control system to solve the problems raised in the above background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] An optimization method for parameters of a steam turbine power-frequency electro-hydraulic control system, the specific steps include:

[0009] Obtain the historical speed curve and power adjustment amount of the steam turbine, add time stamps, divide the speed curve with the time stamps as nodes, the speed curve segment between two adjacent time stamps is a sample, use the speed curve segment as the training set, and the corresponding power adjustment amount as the label to train the convolutional neural network model to obtain a speed regulation prediction model;

[0010] Taking the time stamp when the power adjustment amount was last obtained as the starting point, obtain the real-time speed curve of the steam turbine, and taking the time stamp when the current power adjustment amount is obtained as the end point, form the current speed curve, input the current speed curve into the adjustment prediction model to obtain the predicted power adjustment amount;

[0011] Obtain the conversion coefficient between steam energy and steam turbine power, obtain the energy adjustment amount according to the power adjustment amount, obtain the flow opening formula according to the relationship between the opening of the valve and the flow rate, and obtain the steam energy according to the flow opening formula and the steam pressure data;

[0012] Analyze the formula of steam energy in MATLAB, construct a selection interval according to the steam energy, and obtain the opening interval of the valve under different pressure conditions. Construct an opening score according to the linear representation degree between the opening interval and the selection interval, obtain the minimum opening score, and calibrate the pressure corresponding to the minimum opening score as the control pressure;

[0013] Obtain the actual power adjustment amount, and construct a change index according to the difference between the actual power adjustment amount and the predicted power adjustment amount. Set an index threshold, and judge the pressure required for secondary frequency modulation and the opening interval of the valve according to the relationship between the change index and the index threshold.

[0014] Further, obtain the historical speed curve of the steam turbine, the abscissa of the speed curve is time, the ordinate is speed, obtain the historical power adjustment amount, the power adjustment amount is the power change amount in the secondary frequency modulation process specified by the AGC system, add a time stamp to the power adjustment amount, the time stamp is the time when the AGC system instruction is received, divide the speed curve with the time stamp of the power adjustment amount as the node, the speed curve segment between two adjacent time stamps is a sample, use the divided multiple speed curve segments as the training set, and use the power adjustment amount corresponding to each speed curve segment as the label to input the convolutional neural network model for training, and calibrate the trained model as the adjustment prediction model.

[0015] Further, starting from the timestamp of the last obtained power adjustment amount, the rotational speed curve of the steam turbine is obtained in real time. When the steam turbine triggers primary frequency modulation, taking the time of obtaining the AGC system instruction this time as the end point, the current rotational speed curve is formed. The current rotational speed curve is input into the adjustment prediction model to output the predicted power adjustment amount.

[0016] Further, the conversion coefficient for converting steam energy into the power of the steam turbine is obtained through the merchant. According to the power adjustment amount, the energy adjustment amount of the steam in the steam turbine chamber per unit time is obtained. The basis formula is as follows:

[0017]

[0018] Wherein, is the energy adjustment amount, is the conversion coefficient, is the predicted power adjustment amount;

[0019] The valve of the model used by this steam turbine is experimented in the laboratory to obtain multiple groups of data. Taking the opening of the valve as the independent variable and the flow rate per unit time corresponding to the opening as the dependent variable, the flow rate-opening formula is obtained. The basis formula is as follows:

[0020]

[0021] Wherein, is the flow rate, is the opening, is the relationship function between the flow rate and the opening;

[0022] The steam energy input into the steam turbine chamber per unit time is obtained. The basis formula is as follows:

[0023]

[0024] Wherein, is the steam energy, is the steam pressure, is expressed as the entropy value when the steam pressure is and is obtained through the steam table.

[0025] Further, the formula of the steam energy is analyzed in MATLAB. Taking the steam pressure and the opening as the independent variables and the steam energy as the dependent variable, a surface diagram is constructed. Based on the current steam energy, a selection interval is constructed. The selection interval is the adjustment range of the steam energy, and the range of the selection interval is where is the current steam energy, is the energy adjustment amount;

[0026] The opening intervals under different steam pressure conditions within the selection interval are obtained respectively. The logic for obtaining the opening intervals is:

[0027] For the steam pressure , the corresponding entropy value is obtained through the steam table , and then the opening range under the steam pressure is obtained. The formula is as follows: where,

[0028]

[0029]

[0030] Among them, is the lower limit of the opening range, is the upper limit of the opening range, is the inverse function of the relationship function between the flow rate and the opening.

[0031] Furthermore, according to the linear representation degree between the opening range and the selection range, the opening score under different pressures is constructed. The formula is as follows:

[0032]

[0033] where, is the opening score when the pressure is , is the steam energy, is the opening, is the minimum value within the opening range when the pressure is , is the maximum value within the opening range when the pressure is , is the coefficient;

[0034] The pressure corresponding to the opening score with the largest opening score is calibrated as the control pressure.

[0035] Furthermore, the pressure in the steam chamber of this steam turbine is increased to the control pressure. According to the AGC system, the actual power adjustment amount is obtained. According to the actual power change amount and the predicted power change amount, a change index is formed. The formula is as follows:

[0036]

[0037] where, is the change index, is the predicted power adjustment amount, is the actual power adjustment amount, is a constant.

[0038] Further, an exponential threshold is set. When the change index does not exceed the exponential threshold, secondary frequency modulation is performed within the current control pressure and the opening range of the valve. When the change index exceeds the exponential threshold, the actual power adjustment amount is used to replace the predicted power adjustment amount, the steam energy is regenerated, the opening range and the control pressure are re-obtained in MATLAB, and secondary frequency modulation is performed according to the newly obtained control pressure and opening range.

[0039] Further, in this method, the unit time lengths of the steam energy, the predicted power adjustment amount, and the actual power adjustment amount are the same.

[0040] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0041] The present invention obtains the predicted power adjustment amount by training the speed regulation prediction model, constructs the corresponding formula among the steam energy, the valve opening, and the pressure, obtains the relationship between the steam energy selection range and the valve opening range in MATLAB, and selects the opening range in which the valve opening and the steam energy show the most linear relationship through the opening score, so as to achieve the precise adjustment of the steam energy, improve the precise control of the steam turbine rotor speed, and improve the control performance of the power frequency electro-hydraulic system. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic diagram of the overall method flow of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0043] In order to make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to specific embodiments.

[0044] It should be noted that unless otherwise defined, the technical terms or scientific terms used in the present invention should have the ordinary meanings understood by those of ordinary skill in the field to which the present invention belongs. The "first", "second", and similar terms used in the present invention do not indicate any order, quantity, or importance, but are only used to distinguish different components. The terms such as "comprising" or "including" mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. The terms such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms such as "upper", "lower", "left", "right", etc. are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0045] Embodiment:

[0046] Please refer to Figure 1 , the present invention provides a technical solution:

[0047] An optimization method for parameters of a steam turbine power-frequency electro-hydraulic control system, the specific steps include:

[0048] Step 1: Obtain the historical speed curve and power adjustment amount of the steam turbine, and add time stamps. Divide the speed curve with the time stamps as nodes. The speed curve segment between two adjacent time stamps is a sample. Use the speed curve segments as the training set and the corresponding power adjustment amounts as labels to train a convolutional neural network model to obtain a speed regulation prediction model;

[0049] The said Step 1 includes the following contents:

[0050] Obtain the historical speed curve of the steam turbine, where the abscissa of the speed curve is time and the ordinate is speed. Obtain the historical power adjustment amount, which is the power change amount in the secondary frequency regulation process specified by the AGC system. Add a time stamp to the power adjustment amount, and the time stamp is the time when the AGC system instruction is received. Divide the speed curve with the time stamp of the power adjustment amount as the node. The speed curve segment between two adjacent time stamps is a sample. Use the divided multiple speed curve segments as the training set, and use the power adjustment amount corresponding to each speed curve segment as the label to input into the convolutional neural network model for training. Calibrate the trained model as an adjustment prediction model.

[0051] Set the convolution kernel size of the convolutional neural network model to 3x3, the number of convolution kernels to 64, the number of fully connected layers to 500, and the learning rate to 0.001.

[0052] The speed curve of the secondary frequency regulation event usually shows a single trend in a short time. This mode has significant temporal locality rather than global long-term dependence, which is suitable for the application scenario of the convolutional neural network model. The 3x3 convolution kernel is the most common convolution kernel size in the convolutional neural network, which can effectively capture local change patterns. Compared with larger convolution kernels, the 3x3 convolution kernel can reduce the computational amount. 64 convolution kernels can avoid redundancy while ensuring the model capacity. 500 neurons provide sufficient non-linear fitting ability to encode complex speed-power relationships, and at the same time avoid gradient disappearance or computational redundancy caused by extreme sparsity. 0.001 is the common initial learning rate of adaptive optimizers such as Adam, which is suitable for medium-complexity models. The prediction task of the power adjustment amount requires a fine weight update step size to avoid ignoring subtle temporal pattern differences due to too large a learning rate, and at the same time avoid slow convergence caused by a low learning rate.

[0053] Train a convolutional neural network model by segmenting the time stamps of the historical speed curve and the power adjustment amount. At the same time, strictly align the training data and labels with the AGC command time stamp as the node to avoid data aliasing in different frequency modulation stages, thereby improving the recognition accuracy of the model for secondary frequency modulation events. This step provides a highly robust speed regulation prediction model for subsequent real-time prediction.

[0054] Step 2: Starting from the time stamp of the previously obtained power adjustment amount, obtain the speed curve of the real-time steam turbine. Taking the time stamp of the current power adjustment amount as the end point, form the current speed curve, and input the current speed curve into the adjustment prediction model to obtain the predicted power adjustment amount;

[0055] The said Step 2 includes the following contents:

[0056] Starting from the time stamp of the previously obtained power adjustment amount, obtain the speed curve of the steam turbine in real time. When the steam turbine triggers primary frequency modulation, taking the time of the current AGC system command as the end point, form the current speed curve, and input the current speed curve into the adjustment prediction model to output the predicted power adjustment amount.

[0057] Step 3: Obtain the conversion coefficient between steam energy and steam turbine power, and obtain the energy adjustment amount according to the power adjustment amount. Obtain the flow opening formula according to the relationship between the opening of the valve and the flow rate, and obtain the steam energy according to the flow opening formula and the steam pressure data;

[0058] The said Step 3 includes the following contents:

[0059] Obtain the conversion coefficient for converting steam energy into steam turbine power through the merchant. According to the power adjustment amount, obtain the energy adjustment amount of the steam in the steam turbine chamber per unit time. The formula is as follows:

[0060]

[0061] Where, is the energy adjustment amount, is the conversion coefficient, is the predicted power adjustment amount;

[0062] Construct the energy adjustment amount of the steam turbine chamber through the conversion coefficient. The performance of different steam turbine chambers is different. The higher the conversion coefficient, the less the energy adjustment amount, and the lower the conversion coefficient, the higher the energy adjustment amount.

[0063] Conduct experiments on the valves of the model used in this steam turbine in the laboratory to obtain multiple groups of data. Taking the opening of the valve as the independent variable and the flow rate per unit time corresponding to the opening as the dependent variable, obtain the flow opening formula. The formula is as follows:

[0064]

[0065] Among them, is the flow rate, is the opening degree, is the relationship function between the flow rate and the opening degree;

[0066] As a preferred embodiment, starting from a valve opening degree of 0%, with an increase of 2% as each test point, the flow rate per unit time at different opening degrees is obtained. The polynomial model is selected as the basis for the flow rate - opening degree formula, and the least - squares method is used to solve the coefficients of each term. 80% of the sample set is used as the training set, and the remaining 20% is used as the verification set. When the maximum relative error is set to be less than 2%, the flow rate - opening degree formula is considered to be obtained.

[0067] The steam energy input into the turbine chamber per unit time is obtained according to the following formula:

[0068]

[0069] Among them, is the steam energy, is the steam pressure, represents the entropy value when the steam pressure is and is obtained through the steam table.

[0070] The relationship between the steam energy and the opening degree is constructed through a thermodynamic model. The higher the entropy value, the higher the steam energy passing through at the same opening degree.

[0071] The quantitative relationship between the power adjustment amount and the energy adjustment amount is established through the conversion coefficient. Combining the flow rate - opening degree formula and the steam energy formula calibrated in the laboratory, the abstract power demand is converted into operable valve opening degree and pressure parameters. This step circumvents the dependence on complex thermodynamic equations through energy - power decoupling and valve characteristic quantification, and at the same time constructs an energy closed - loop model to ensure the real - time matching of steam energy supply and dynamic working conditions (such as pressure fluctuations), providing a physically implementable parameter basis for subsequent optimization.

[0072] Step 4: Analyze the formula of steam energy in MATLAB, construct a selection interval according to the steam energy, and obtain the opening degree interval of the valve under different pressure conditions. Construct an opening degree score according to the linear representation degree between the opening degree interval and the selection interval, obtain the minimum opening degree score, and calibrate the pressure corresponding to the minimum opening degree score as the control pressure;

[0073] The said Step 4 includes the following contents:

[0074] Step 401: Analyze the formula of steam energy in MATLAB, construct a surface graph with steam pressure and opening degree as independent variables and steam energy as the dependent variable, and construct a selection interval based on the current steam energy. The selection interval is the adjustment range of steam energy, and the range of the selection interval is , where is the current steam energy, and is the energy adjustment amount;

[0075] Obtain the opening intervals under different steam pressure conditions within the selection interval respectively. The logic for obtaining the opening intervals is as follows:

[0076] For the steam pressure , obtain the corresponding entropy value through the steam table , and then obtain the opening interval under the steam pressure . The formula based on is:

[0077]

[0078]

[0079] where, is the lower limit of the opening interval, is the upper limit of the opening interval, is the inverse function of the relationship function between the flow rate and the opening.

[0080] Construct a surface plot of steam energy versus pressure and opening in MATLAB, define the selection interval, and at the same time expand the opening interval to 80% - 120% of the domain range. By visualizing the global distribution of steam energy supply, the physical boundary of steam energy adjustment can be clarified, avoiding local optimization traps; expanding the opening interval reserves a judgment space for the inconsistency between the actual power adjustment amount and the predicted power adjustment amount in the later stage, enhancing the tolerance of the system, and ensuring the robustness of parameter optimization.

[0081] Step 402: Construct the opening score under different pressures according to the linear representation degree between the opening interval and the selection interval. The formula based on is as follows:

[0082]

[0083] where, is the opening score when the pressure is , is the steam energy, is the opening, is the minimum value within the opening interval when the pressure is , is the maximum value within the opening interval when the pressure is , is the coefficient;

[0084] The specific value of needs to be determined according to the specific valve system and pressure conditions, and is obtained through experimental data or simulation calculations within the opening interval The variation law to determine the value so that the opening degree scoring value range is .

[0085] Among them, the first partial derivative of the steam energy with respect to the opening degree reflects the change rate of the steam energy with respect to the opening degree, and judges the degree to which the relationship between the steam energy and the opening degree tends to be linear. The second partial derivative of the steam energy with respect to the opening degree reflects the change of the change rate of the steam energy with respect to the opening degree, and judges the linear slope of the steam energy with respect to the opening degree. The smaller the linear slope, the less the steam energy increases when the valve opening increases by the same amount, which reflects that the valve controls the steam energy more precisely. The final result is that the opening degree score is larger. The absolute value is used to avoid negative values caused by the second partial derivative. The smaller the area formed by the curve of the second partial derivative and the horizontal axis within the domain interval, whether the second partial derivative is positive or negative, will make the opening degree score larger. The linear slopes within the opening degree interval are accumulated in the form of integration to judge the control precision degree from the entire opening degree interval.

[0086] Select the pressure at which the opening degree score is the largest as the control pressure.

[0087] According to the most accurate opening degree interval, obtain the optimal control pressure, which can quantify the curvature change of the energy-opening curve, and preferentially select the region with high linearity to avoid control oscillation or instability caused by the non-linear region, so as to achieve the optimal balance between the regulation efficiency and the system stability.

[0088] Step 5: Obtain the actual power adjustment amount, and construct a change index according to the difference between the actual power adjustment amount and the predicted power adjustment amount, set an index threshold, and judge the pressure required for secondary frequency modulation and the opening degree interval of the valve according to the relationship between the change index and the index threshold.

[0089] The said Step 5 includes the following contents:

[0090] Step 501: Raise the pressure of the steam chamber of this steam turbine to the control pressure, obtain the actual power adjustment amount according to the AGC system, and construct a change index according to the actual power change amount and the predicted power change amount. The formula is as follows:

[0091]

[0092] Among them, is the change index, is the predicted power adjustment amount, is the actual power adjustment amount, is a constant.

[0093] It is the baseline value of the change index, which is set empirically based on historical operating conditions and combines with the index threshold to achieve the sensitivity of feedback control.

[0094] A change index is constructed based on the difference between the predicted and actual power adjustment amounts. The sensitivity of the prediction deviation can be amplified through an exponential function. When the actual power adjustment amount exceeds the predicted power adjustment amount, the change index will increase. Moreover, the higher the actual power adjustment amount exceeds the predicted power adjustment amount, the faster the change index increases, providing a quantitative guarantee for whether the prediction result can meet the actual situation.

[0095] Step 502: Set the index threshold. When the change index does not exceed the index threshold, perform secondary frequency modulation within the current control pressure and valve opening range. When the change index exceeds the index threshold, replace the predicted power adjustment amount with the actual power adjustment amount, regenerate the steam energy, re-obtain the opening range and control pressure in MATLAB, and perform secondary frequency modulation according to the newly obtained control pressure and opening range.

[0096] Setting the index threshold for feedback control - within the threshold, continue to perform secondary frequency modulation using the current parameters. When exceeding the threshold, regenerate the steam energy and optimize the control pressure and opening range, which can achieve dual-mode adaptive control: maintain high-efficiency frequency modulation under steady-state conditions, quickly correct parameters under transient or abnormal conditions, form a "prediction - execution - feedback" closed loop, significantly improve the system's adaptability and fault tolerance to sudden disturbances, and ultimately ensure the global frequency modulation accuracy and stability.

[0097] In this embodiment, the unit time lengths of the steam energy, predicted power adjustment amount, and actual power adjustment amount in this method are the same.

[0098] The above formulas are all dimensionless and take their numerical calculations. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the real situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0099] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0100] The unit described as a separation component may or may not be physically separated. The component displayed as a unit may or may not be a physical unit. It may be located in one place or distributed across multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0101] As described above, only the specific implementation manners of this application are provided, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed by this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application.

Claims

1. An optimization method for parameters of a steam turbine power-frequency electro-hydraulic control system, characterized in that, The specific steps include: Obtain the historical speed curve and power adjustment amount of the steam turbine, add timestamps, divide the speed curve with timestamps as nodes, and use the speed curve segment between two adjacent timestamps as a sample. Use the speed curve segment as the training set and the corresponding power adjustment amount as the label to train the convolutional neural network model to obtain the speed control prediction model; Taking the timestamp of the last obtained power adjustment amount as the starting point, obtain the real-time speed curve of the steam turbine, and form the current speed curve with the timestamp of obtaining the current power adjustment amount as the end point. Input the current speed curve into the adjustment prediction model to obtain the predicted power adjustment amount; Obtain the conversion coefficient between steam energy and steam turbine power, obtain the energy adjustment amount according to the predicted power adjustment amount, obtain the flow opening formula based on the relationship between the valve opening and flow rate, and obtain the steam energy according to the flow opening formula and steam pressure data; Analyze the steam energy formula in MATLAB, construct a selection interval based on the steam energy, and obtain the opening interval of the valve under different steam pressure conditions. Construct an opening score based on the linear representation degree between the opening interval and the selection interval, obtain the minimum opening score, and calibrate the pressure corresponding to the minimum opening score as the control pressure; Obtain the actual power adjustment amount, construct a change index according to the difference between the actual power adjustment amount and the predicted power adjustment amount, set an index threshold, and judge the pressure and valve opening interval required for secondary frequency modulation according to the relationship between the change index and the index threshold.

2. The optimization method for the parameters of a steam turbine power-frequency electro-hydraulic control system according to claim 1, wherein: Obtain the historical speed curve of the steam turbine, where the abscissa of the speed curve is time and the ordinate is speed. Obtain the historical power adjustment amount, where the power adjustment amount is the power change amount during the secondary frequency modulation specified by the AGC system. Add a timestamp to the power adjustment amount, where the timestamp is the time when the AGC system instruction is received. Divide the speed curve with the timestamp of the power adjustment amount as the node, and use the speed curve segment between two adjacent timestamps as a sample. Use the divided multiple speed curve segments as the training set and input the power adjustment amount corresponding to each speed curve segment as the label into the convolutional neural network model for training. Calibrate the trained model as the adjustment prediction model.

3. An optimization method for parameters of a steam turbine power-frequency electro-hydraulic control system according to claim 2, characterized in that: Taking the timestamp of the last obtained power adjustment amount as the starting point, obtain the real-time speed curve of the steam turbine. When the steam turbine triggers primary frequency modulation, form the current speed curve with the time of obtaining the AGC system instruction this time as the end point. Input the current speed curve into the adjustment prediction model to output the predicted power adjustment amount.

4. An optimization method for parameters of a steam turbine power-frequency electro-hydraulic control system according to claim 3, characterized in that: Obtain the conversion coefficient of converting steam energy to steam turbine power through the merchant, and obtain the energy adjustment amount of the steam in the steam turbine chamber per unit time according to the predicted power adjustment amount. The formula is as follows: ， wherein, is the energy adjustment amount, is the conversion coefficient, is the predicted power adjustment amount; Conduct experiments on the valves of this type of steam turbine in the laboratory to obtain multiple groups of data. Use the valve opening as the independent variable and the flow rate per unit time corresponding to the opening as the dependent variable to obtain the flow opening formula. The formula is as follows: ， Among them, is the flow rate, is the opening degree, is the relationship function between the flow rate and the opening degree; Obtain the steam energy input into the steam turbine chamber per unit time. The formula is as follows: ， Among them, is the steam energy, is the steam pressure, is expressed as the entropy value when the steam pressure is and is obtained through the steam table.

5. An optimization method for parameters of a steam turbine power-frequency electro-hydraulic control system according to claim 4, characterized in that: Analyze the formula of steam energy in MATLAB, construct a surface plot with steam pressure and opening as independent variables and steam energy as the dependent variable, and construct a selection interval based on the current steam energy. The selection interval is the adjustment range of steam energy, and the range of the selection interval is , where is the current steam energy, is the energy adjustment amount; Obtain the opening intervals under different steam pressure conditions within the selection interval respectively. The logic for obtaining the opening interval is: For the steam pressure , obtain the corresponding entropy value through the steam table , and then obtain the opening range under the steam pressure . The formula is as follows: ， ， Among them, is the lower limit of the opening range, is the upper limit of the opening range, is the inverse function of the relationship function between the flow rate and the opening.

6. The optimization method for parameters of a steam turbine power-frequency electro-hydraulic control system according to claim 5, characterized in that: Construct the opening degree score under different pressures according to the linear representation degree of the opening degree interval and the selection interval, and the basis formula is as follows: ， Wherein, is the opening degree score when the pressure is ; is the steam energy; is the opening degree; is the minimum value within the opening degree range when the pressure is ; is the maximum value within the opening degree range when the pressure is ; is the coefficient. Calibrate the pressure at which the opening degree score with the largest opening degree score is selected as the control pressure.

7. An optimization method for parameters of a steam turbine power-frequency electro-hydraulic control system according to claim 6, characterized in that: Raise the pressure of the steam chamber of this steam turbine to the control pressure, obtain the actual power adjustment amount according to the AGC system, and constitute a change index according to the actual power change amount and the predicted power change amount. The basis formula is as follows: ， Among them, is the change index, is the predicted power adjustment amount, is the actual power adjustment amount, is a constant.

8. An optimization method for parameters of a steam turbine power-frequency electro-hydraulic control system according to claim 7, characterized in that: Set an index threshold. When the change index does not exceed the index threshold, perform secondary frequency modulation in the current control pressure and the opening degree interval of the valve. When the change index exceeds the index threshold, replace the predicted power adjustment amount with the actual power adjustment amount, regenerate the steam energy, re-obtain the opening degree interval and the control pressure in MATLAB, and perform secondary frequency modulation according to the newly obtained control pressure and opening degree interval.

9. An optimization method for parameters of a steam turbine power-frequency electro-hydraulic control system according to claim 8, characterized in that: In this method, the unit time lengths of the steam energy, the predicted power adjustment amount, and the actual power adjustment amount are the same.

Images