Regression analysis method and system for temperature and power of combustion chamber of combined cycle system of gas turbine

Abstract

The invention relates to the technical field of gas turbine combined cycle systems, in particular to a gas turbine combined cycle system combustion chamber temperature and power regression analysis method and system. Establishing a regression model taking the temperature data of each temperature measurement position as an input variable and the output power as an output variable; adjusting the weight coefficient of each item in the regression model through the temperature features; and predicting the output power of the gas turbine combined cycle system by using the regression model. According to the method, by calculating the standard deviation, the skewness coefficient and the kurtosis coefficient, the dispersion degree, symmetry and sharp degree of the combustion chamber temperature are comprehensively reflected, a basis is provided for weight adjustment, a regression model can better reflect the dynamic relation between the combustion chamber temperature and power, importance of the temperature of each temperature measurement position in power prediction can be reasonably reflected, and power prediction accuracy is improved. And the accuracy of output power prediction is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of gas turbine combined cycle systems, and in particular to a regression analysis method and system for combustion chamber temperature and power of a gas turbine combined cycle system. Background Art

[0002] With their dual advantages of high efficiency and environmental friendliness, gas turbine combined cycle systems have become key technological equipment in the modern energy sector, playing an indispensable role in power generation and industrial power supply. As the core component of the system, the combustion chamber couples multiple physical fields, including turbulent combustion, chemical reactions, radiation heat transfer, and convection heat transfer, resulting in a highly non-uniform temperature distribution. This temperature distribution and variations directly affect the operating efficiency, emissions performance, and equipment life of the gas turbine. The temperature field inside the combustion chamber is highly non-uniform and dynamically time-varying, and the temperature distribution patterns vary significantly under different load conditions. Therefore, accurately predicting the system output power using combustion chamber temperature data is of great significance for optimizing operational control, improving energy utilization, and reducing maintenance costs. Summary of the Invention

[0003] (1) Purpose of the invention

[0004] The purpose of the present invention is to provide a regression analysis method and system for combustion chamber temperature and power of a gas turbine combined cycle system, which predicts output power by adjusting the weight coefficient of each temperature measurement position through temperature characteristics, thereby improving the accuracy and reliability of power prediction.

[0005] (2) Technical solution

[0006] To solve the above problems, the present invention provides a regression analysis method for combustion chamber temperature and power of a gas turbine combined cycle system, comprising:

[0007] Calculate temperature characteristics based on the temperature data of each temperature measurement location, wherein the temperature characteristics include: standard deviation index, skewness coefficient and kurtosis coefficient;

[0008] Establish a regression model with the temperature data of each temperature measurement location as the input variable and the output power as the output variable;

[0009] Adjusting the weight coefficients of each item in the regression model according to the temperature characteristics;

[0010] The regression model is used to predict the output power of a gas turbine combined cycle system.

[0011] In another aspect of the present invention, preferably, calculating the temperature characteristics based on the temperature data of each temperature measurement position includes:

[0012] The interpolation algorithm is used to perform spatial interpolation on the temperature data according to the location of each measuring point to obtain wide-area temperature data;

[0013] The standard deviation coefficient, the skewness coefficient, and the kurtosis coefficient are calculated based on the wide-area temperature data.

[0014] In another aspect of the present invention, preferably, the adopting of an interpolation algorithm to spatially interpolate the temperature data according to the positions of each measuring point to obtain wide-area temperature data includes:

[0015] According to the geometric structure of the combustion chamber and the three-dimensional spatial coordinates of each temperature measurement position, an interpolation interval starting from each temperature measurement position is determined;

[0016] According to the temperature data and interpolation interval of each measuring point, the temperature of the internal space of the combustion chamber is interpolated based on the temperature gradient to obtain the wide-area temperature data of the combustion chamber.

[0017] In another aspect of the present invention, preferably, the calculation of the standard deviation coefficient includes:

[0018] Calculating a temperature average of all wide-area temperature data based on the wide-area temperature data;

[0019] Calculating the standard deviation of the wide-area temperature data within the combustion chamber based on the temperature average value;

[0020] The ratio of the standard deviation to the temperature average is calculated, where the ratio is the standard deviation coefficient.

[0021] In another aspect of the present invention, preferably,

[0022] The calculation of the skewness coefficient includes:

[0023] Calculating a temperature average and a standard deviation of the wide-area temperature data based on the wide-area temperature data;

[0024] Calculate the standardized value of each wide-area temperature data based on the temperature data, average value and standard deviation of the measuring points;

[0025] The skewness coefficient is calculated based on the standardized values ​​of all wide-area temperature data.

[0026] In another aspect of the present invention, preferably,

[0027] The calculation of the kurtosis coefficient includes:

[0028] Calculating a temperature average of all wide-area temperature data based on the wide-area temperature data;

[0029] Obtaining a fourth-order central moment according to the temperature average value and each wide-area temperature data;

[0030] The kurtosis coefficient is obtained based on the fourth-order central moment and the standard deviation of the wide-area temperature data.

[0031] In another aspect of the present invention, preferably, the regression model is based on a linear regression model;

[0032] Each temperature measurement location, the corresponding temperature data and the weight coefficient are taken as a feature vector;

[0033] The characteristic vector is the input variable of the regression model, and the output power is the output variable.

[0034] In another aspect of the present invention, preferably, adjusting the weight coefficients of each item in the regression model according to the temperature characteristics includes:

[0035] Establishing a lookup table, wherein the lookup table includes a correspondence between temperature characteristics and weight coefficients of each temperature measurement position;

[0036] According to the lookup table and the temperature characteristics, the weight coefficient of each temperature measurement position is obtained.

[0037] In another aspect of the present invention, preferably, said establishing a lookup table comprises:

[0038] Based on the historical temperature data, the standard deviation coefficient, skewness coefficient and kurtosis coefficient are calculated, and clustering is performed to divide multiple temperature characteristic intervals. Each temperature characteristic interval is associated with a set of weight coefficients.

[0039] The weight coefficient associated with each temperature characteristic interval is solved by the particle swarm optimization algorithm based on the historical temperature data and the corresponding output power, with the goal of minimizing the power prediction error.

[0040] In another aspect of the present invention, a regression analysis method for combustion chamber temperature and power of a gas turbine combined cycle system is preferably provided, comprising:

[0041] Calculation module: calculates temperature characteristics based on the temperature data of each temperature measurement location, and the temperature characteristics include: standard deviation index, skewness coefficient and kurtosis coefficient;

[0042] Establishing module: establishing a regression model with the temperature data of each temperature measurement location as input variable and output power as output variable;

[0043] Adjustment module: adjusts the weight coefficients of each item in the regression model according to the temperature characteristics;

[0044] Prediction module: using the regression model to predict the output power of the gas turbine combined cycle system.

[0045] (3) Beneficial effects

[0046] The above technical solution of the present invention has the following beneficial technical effects:

[0047] The present invention comprehensively reflects the discreteness, symmetry and sharpness of the combustion chamber temperature by calculating the standard deviation, skewness coefficient and kurtosis coefficient, provides a basis for weight adjustment, better reflects the dynamic relationship between the combustion chamber temperature and power, helps to reasonably reflect the importance of the temperature at each temperature measurement position in power prediction, and improves the accuracy of output power prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is an overall flow chart of an embodiment of the present invention. DETAILED DESCRIPTION

[0049] To make the objectives, technical solutions, and advantages of the present invention more clearly understood, the present invention will be further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings. It should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present invention. In addition, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present invention.

[0050] Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0051] In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0052] The present invention will be described in more detail below with reference to the accompanying drawings. In each of the accompanying drawings, identical elements are represented by similar reference numerals. For the sake of clarity, the various parts in the accompanying drawings are not drawn to scale.

[0053] Example 1

[0054] A regression analysis method for combustion chamber temperature and power in a gas turbine combined cycle system. Figure 1 FIG. 1 shows an overall flow chart of an embodiment of the present invention, as shown in FIG. Figure 1 Shown, including:

[0055] Temperature characteristics are calculated based on the temperature data from each temperature measurement location. These temperature characteristics include standard deviation, skewness, and kurtosis. The temperature distribution at different locations within the combustion chamber is complex and dynamically changing, making a single temperature data point inadequate for fully reflecting the temperature state. Therefore, in this embodiment, the collected temperature data is from various locations within the combustion chamber. The specific location of each temperature measurement location within the combustion chamber is not limited here; the temperature measurement locations can be evenly or unevenly distributed. Temperature data can be acquired using thermocouples, for example. The standard deviation reflects the degree of dispersion in the temperature data. A larger standard deviation indicates greater temperature fluctuations at each measurement point, implying less uniform temperature distribution within the combustion chamber. For example, when a gas turbine is operating at high load, a large standard deviation may indicate the presence of localized high-temperature regions, which can affect combustion efficiency. The skewness describes the degree of asymmetry in the temperature data distribution. A positive skewness indicates a longer right-hand tail of the temperature data distribution, indicating the presence of high extreme temperature values. A negative skewness indicates a longer left-hand tail, indicating a higher number of extreme low-temperature values. The skewness can be used to determine whether there are abnormally high or low-temperature regions within the combustion chamber. The kurtosis coefficient measures the sharpness of the peak in the temperature data distribution. A higher kurtosis coefficient indicates that the data distribution is more concentrated around the mean, and the probability of extreme values ​​is relatively low. Conversely, a lower kurtosis coefficient indicates that the data distribution is more dispersed, and extreme temperatures are more likely to occur.

[0056] Furthermore, in this embodiment, the temperature characteristics are calculated based on the temperature data of each temperature measurement position, including:

[0057] An interpolation algorithm is used to perform spatial interpolation on the temperature data according to the location of each measuring point to obtain wide-area temperature data. In the actual monitoring process, due to limitations of cost, technology and other factors, temperature measuring equipment can often only be arranged at limited measuring point locations, and discrete measuring point data cannot intuitively reflect the temperature distribution of the entire monitoring area. The introduction of the spatial interpolation algorithm is to reasonably expand these discrete temperature data and generate continuous wide-area temperature data, thereby more comprehensively describing the temperature field of the entire area. Interpolation algorithms include inverse distance weighted interpolation, Kriging interpolation, spline interpolation, etc. In this embodiment, the interpolation algorithm is used to perform spatial interpolation on the temperature data according to the location of each measuring point to obtain wide-area temperature data, including:

[0058] Based on the geometric structure of the combustion chamber and the three-dimensional spatial coordinates of each temperature measurement position, the interpolation interval starting from each temperature measurement position is determined; the gas turbine combustion chamber is annular in structure and is equipped with multiple fuel injectors and air inlets. When arranging the temperature measurement points, temperature sensors are installed at different radial, axial, and circumferential positions. When determining the interpolation interval, based on the geometric boundaries of the combustion chamber and taking each temperature measurement position as the starting point, a reasonable interpolation range is defined in combination with the structural characteristics of the combustion chamber. For example, for temperature measurement points near the center of the flame, due to the drastic temperature changes, the interpolation interval may be relatively small to more accurately capture the rapid temperature changes; in areas with relatively uniform temperature distribution, the interpolation interval can be appropriately expanded to reduce unnecessary calculations.

[0059] Based on the temperature data at each measuring point and the interpolation interval, the temperature inside the combustion chamber is interpolated based on the temperature gradient to obtain the wide-area temperature data for the combustion chamber. The temperature gradient reflects the rate of change of temperature over space. Interpolation based on the temperature gradient can fully utilize the spatial variation of temperature data, making the interpolation result more accurate. After determining the interpolation interval, for each interpolated point, the temperature gradient between it and the adjacent measuring points is first calculated. In actual calculations, the temperature gradients in three directions are comprehensively considered, and the overall temperature gradient at that point is obtained through methods such as vector synthesis. After the temperature gradient is obtained, the interpolation formula is used to calculate the temperature gradient based on the interpolation interval and the location of the interpolated point. For example, linear interpolation can be used to calculate the temperature estimate of the interpolated point proportionally to the distance between the interpolated point and the known measuring point along the direction of the temperature gradient. By performing this calculation for each interpolated point, the temperature data is gradually filled in for the entire combustion chamber interior, ultimately obtaining complete wide-area temperature data. The wide-area temperature data can be presented as a three-dimensional temperature cloud map, which intuitively displays the temperature distribution within the combustion chamber.

[0060] According to the wide-area temperature data, the standard deviation coefficient, the skewness coefficient and the kurtosis coefficient are calculated. The specific content of calculating the standard deviation coefficient, the skewness coefficient and the kurtosis coefficient is not limited here. In this embodiment,

[0061] The calculation of the standard deviation coefficient includes:

[0062] Based on the wide-area temperature data, the temperature average of all wide-area temperature data is calculated; based on the combustion chamber wide-area temperature data, the average is calculated by summing and dividing by the total amount of data, which can reflect the overall level of the temperature inside the combustion chamber.

[0063] Based on the temperature average, the standard deviation of the wide-area temperature data in the combustion chamber is calculated; the standard deviation is used to measure the degree of dispersion of the data relative to the average, and in the combustion chamber temperature analysis, it intuitively reflects the fluctuation of the temperature data.

[0064] The temperature mean and standard deviation are calculated using the following formula:

[0065]

[0066] in, represents the average temperature, σ represents the standard deviation of temperature, m represents the number of wide-area temperature data, j represents the number of wide-area temperature data, T j Represents the j-th wide-area temperature data.

[0067] The ratio of the standard deviation to the temperature average is calculated, where the ratio is the standard deviation coefficient. The standard deviation coefficient is expressed using the following formula:

[0068]

[0069] Where C represents the standard deviation coefficient. The standard deviation reflects the degree of dispersion of the temperature data. A larger standard deviation indicates more drastic temperature fluctuations at each measurement point, which means that the temperature distribution in the combustion chamber is less uniform.

[0070] The calculation of the skewness coefficient includes:

[0071] Calculating a temperature average and a standard deviation of the wide-area temperature data based on the wide-area temperature data;

[0072] The standardized value of each wide-area temperature data is calculated based on the temperature data of the measuring points, the average value, and the standard deviation. In this embodiment, the standardized value of each wide-area temperature data is calculated using the following formula:

[0073]

[0074] The skewness coefficient is calculated based on the normalized values ​​of all wide-area temperature data using the following formula:

[0075]

[0076] Among them, S represents the skewness coefficient, x j The skewness coefficient indicates the normalized value of the j-th wide-area temperature data. The skewness coefficient can be used to determine whether there are abnormally high or low temperature areas in the combustion chamber.

[0077] The calculation of the kurtosis coefficient includes:

[0078] Calculating a temperature average of all wide-area temperature data based on the wide-area temperature data;

[0079] Obtaining a fourth-order central moment according to the temperature average value and each wide-area temperature data;

[0080] The fourth-order central moment is calculated using the following formula:

[0081]

[0082] The kurtosis coefficient is obtained based on the fourth-order central moment and the standard deviation of the wide-area temperature data.

[0083] The kurtosis coefficient is calculated using the following formula:

[0084]

[0085] Where K represents the kurtosis coefficient and μ4 represents the fourth-order central moment. If the kurtosis coefficient is greater than zero, the wide-area temperature data distribution is more peaked than the normal distribution; if the kurtosis coefficient is less than zero, the wide-area temperature data distribution is flatter than the normal distribution; if the kurtosis coefficient is close to zero, the wide-area temperature data distribution is close to the normal distribution.

[0086] A regression model is established with the temperature data at each temperature measurement location as the input variable and the output power as the output variable. Traditional regression models often assume that the relationship between variables is fixed. However, in a gas turbine combined cycle system, the operating state of the combustion chamber will continuously change with factors such as time and load, and the relationship between temperature and power is not static. Therefore, in this embodiment, the weight coefficients of each item in the regression model are adjusted based on the temperature characteristics. The degree of influence of different temperature measurement locations on the system output power is not the same, and this degree of influence will also change with changes in temperature characteristics. In this embodiment, the regression model is based on a linear regression model.

[0087] Each temperature measurement location, its corresponding temperature data, and its weight coefficient are considered a feature vector. Based on the number of temperature measurement locations, a number of feature vectors are generated. These feature vectors serve as input variables for the regression model, with output power serving as the output variable. The feature vectors, as input variables for the regression model, also include characteristics such as pressure and flow rate. Model parameters related to these characteristics can be obtained through training based on historical data. Temperature measurement locations can be set at preset intervals throughout the combustion chamber, or they can be located in the combustion chamber's inlet, flame, or mixing zones. The inlet zone temperature reflects the degree of fuel preheating and the initial temperature of the mixture. A higher inlet temperature facilitates greater heat release per unit time, thereby increasing output power. The flame zone temperature is the most direct indicator of combustion intensity. A higher flame temperature means more heat is released per unit time and per unit volume. The mixing zone is the temperature of the fuel after the secondary air or dilution air is injected downstream of the flame, resulting in a cooling of the gas mixture. It directly determines the enthalpy of the working fluid upon entry into the turbine. A higher turbine inlet temperature results in greater power output.

[0088] In this embodiment, the weight coefficients of each item in the regression model are adjusted according to the temperature characteristics, including:

[0089] A lookup table is established, which includes a corresponding relationship between temperature characteristics and weight coefficients of each temperature measurement position; further, in the embodiment, the establishment of the lookup table includes:

[0090] According to historical temperature data, a standard deviation coefficient, a skewness coefficient, and a kurtosis coefficient are calculated, the standard deviation coefficient, the skewness coefficient, and the kurtosis coefficient obtained by calculation based on the historical temperature data are spliced into a three-dimensional feature vector, a plurality of three-dimensional feature vectors are clustered, an unsupervised learning algorithm is used to cluster the feature vectors, similar temperature distribution modes are divided into the same interval, and a plurality of temperature characteristic intervals are divided, each temperature characteristic interval is associated with a group of weight coefficients;

[0091] The weight coefficients associated with each temperature characteristic interval are solved by a particle swarm optimization algorithm based on historical temperature data and corresponding output power, with the objective of minimizing power prediction error. The optimization objective of the weight coefficients is to minimize the power prediction error, that is, by adjusting the weight of each temperature measurement point, the deviation between the output power of the model and the actual power is minimized.

[0092] According to the lookup table and the temperature characteristics, the weight coefficients of each temperature measurement position are obtained. By using different weight coefficients, the accuracy of prediction is improved.

[0093] The regression model is used to predict the output power of the gas turbine combined cycle system.

[0094] The present application comprehensively reflects the dispersion degree, symmetry and sharpness of the combustion chamber temperature by calculating the standard deviation, skewness coefficient and kurtosis coefficient, provides a basis for weight adjustment, better reflects the dynamic relationship between the combustion chamber temperature and the power, and helps to reasonably reflect the importance of the temperature at each temperature measurement position in power prediction, and improves the accuracy of output power prediction.

[0095] Embodiment two

[0096] A regression analysis method for the combustion chamber temperature and power of a gas turbine combined cycle system, comprising:

[0097] The calculation module calculates temperature characteristics according to temperature data of each temperature measurement position, the temperature characteristics including a standard deviation index, a skewness coefficient, and a kurtosis coefficient;

[0098] The establishment module establishes a regression model with temperature data of each temperature measurement position as an input variable and output power as an output variable;

[0099] The adjustment module adjusts the weight coefficients of each item in the regression model through the temperature characteristics;

[0100] The prediction module uses the regression model to predict the output power of the gas turbine combined cycle system.

[0101] It should be understood that the above-described specific embodiments of the present invention are merely illustrative or illustrative of the principles of the present invention and do not constitute limitations of the present invention. Therefore, any modifications, equivalent substitutions, improvements, etc. made without departing from the spirit and scope of the present invention should be included within the scope of protection of the present invention. In addition, the appended claims are intended to cover all variations and modifications that fall within the scope and metes and bounds of the appended claims, or equivalents thereof.

[0102] The present invention has been described above with reference to the embodiments thereof. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of the present invention. The scope of the present invention is defined by the appended claims and their equivalents. Those skilled in the art may make various substitutions and modifications without departing from the scope of the present invention, and such substitutions and modifications are intended to fall within the scope of the present invention.

[0103] Although the embodiments of the present invention have been described in detail, it should be understood that the various changes, substitutions, and alterations could be made hereto without departing from the spirit and scope of the invention.

[0104] Obviously, the above embodiments are merely examples for clarity of explanation and are not intended to limit the implementation methods. Those skilled in the art will readily appreciate that other variations or modifications based on the above descriptions are possible. It is not necessary and impossible to enumerate all implementation methods here. Obvious variations or modifications arising therefrom remain within the scope of protection of the present invention.

Claims

1. A regression analysis method for combustion chamber temperature and power of a gas turbine combined cycle system, characterized in that: include: Calculate temperature characteristics based on the temperature data of each temperature measurement location, wherein the temperature characteristics include: standard deviation index, skewness coefficient and kurtosis coefficient; Establish a regression model with the temperature data of each temperature measurement location as the input variable and the output power as the output variable; Adjusting the weight coefficients of each item in the regression model according to the temperature characteristics; The regression model is used to predict the output power of a gas turbine combined cycle system.

2. The regression analysis method of combustion chamber temperature and power of a gas turbine combined cycle system according to claim 1, characterized in that: Calculate the temperature characteristics based on the temperature data at each temperature measurement location, including: The interpolation algorithm is used to perform spatial interpolation on the temperature data according to the location of each measuring point to obtain wide-area temperature data; The standard deviation coefficient, the skewness coefficient, and the kurtosis coefficient are calculated based on the wide-area temperature data.

3. The regression analysis method of combustion chamber temperature and power of a gas turbine combined cycle system according to claim 1, characterized in that: The interpolation algorithm is used to spatially interpolate the temperature data according to the position of each measuring point to obtain wide-area temperature data, including: According to the geometric structure of the combustion chamber and the three-dimensional spatial coordinates of each temperature measurement position, an interpolation interval starting from each temperature measurement position is determined; According to the temperature data and interpolation interval of each measuring point, the temperature of the internal space of the combustion chamber is interpolated based on the temperature gradient to obtain the wide-area temperature data of the combustion chamber.

4. The regression analysis method of combustion chamber temperature and power of a gas turbine combined cycle system according to claim 2, characterized in that: The calculation of the standard deviation coefficient includes: Calculating a temperature average of all wide-area temperature data based on the wide-area temperature data; Calculating the standard deviation of the wide-area temperature data within the combustion chamber based on the temperature average value; The ratio of the standard deviation to the temperature average is calculated, where the ratio is the standard deviation coefficient.

5. The regression analysis method of combustion chamber temperature and power of a gas turbine combined cycle system according to claim 2, characterized in that: The calculation of the skewness coefficient includes: Calculating a temperature average and a standard deviation of the wide-area temperature data based on the wide-area temperature data; Calculate the standardized value of each wide-area temperature data based on the temperature data, average value and standard deviation of the measuring points; The skewness coefficient is calculated based on the standardized values ​​of all wide-area temperature data.

6. The regression analysis method of combustion chamber temperature and power of a gas turbine combined cycle system according to claim 2, characterized in that: The calculation of the kurtosis coefficient includes: Calculating a temperature average of all wide-area temperature data based on the wide-area temperature data; Obtaining a fourth-order central moment according to the temperature average value and each wide-area temperature data; The kurtosis coefficient is obtained based on the fourth-order central moment and the standard deviation of the wide-area temperature data.

7. The regression analysis method of combustion chamber temperature and power of a gas turbine combined cycle system according to claim 1, characterized in that: The regression model is based on a linear regression model; Each temperature measurement location, the corresponding temperature data and the weight coefficient are taken as a feature vector; The characteristic vector is the input variable of the regression model, and the output power is the output variable.

8. The regression analysis method for combustion chamber temperature and power of a gas turbine combined cycle system according to claim 2, characterized in that: Adjusting the weight coefficients of each item in the regression model according to the temperature characteristics includes: Establishing a lookup table, wherein the lookup table includes a correspondence between temperature characteristics and weight coefficients of each temperature measurement position; According to the lookup table and the temperature characteristics, the weight coefficient of each temperature measurement position is obtained.

9. The regression analysis method for combustion chamber temperature and power of a gas turbine combined cycle system according to claim 7, characterized in that: The establishing of the lookup table comprises: Based on the historical temperature data, the standard deviation coefficient, skewness coefficient and kurtosis coefficient are calculated, and clustering is performed to divide multiple temperature characteristic intervals. Each temperature characteristic interval is associated with a set of weight coefficients. The weight coefficient associated with each temperature characteristic interval is solved by the particle swarm optimization algorithm based on the historical temperature data and the corresponding output power, with the goal of minimizing the power prediction error.

10. A regression analysis method for combustion chamber temperature and power of a gas turbine combined cycle system, characterized in that: include: Calculation module: calculates temperature characteristics based on the temperature data of each temperature measurement location, and the temperature characteristics include: standard deviation index, skewness coefficient and kurtosis coefficient; Establishing module: establishing a regression model with the temperature data of each temperature measurement location as input variable and output power as output variable; Adjustment module: adjusts the weight coefficients of each item in the regression model according to the temperature characteristics; Prediction module: using the regression model to predict the output power of the gas turbine combined cycle system.

Images