CN113111456A - Online interval identification method for key operating parameters of gas turbine - Google Patents
Online interval identification method for key operating parameters of gas turbine Download PDFInfo
- Publication number
- CN113111456A CN113111456A CN202110372207.1A CN202110372207A CN113111456A CN 113111456 A CN113111456 A CN 113111456A CN 202110372207 A CN202110372207 A CN 202110372207A CN 113111456 A CN113111456 A CN 113111456A
- Authority
- CN
- China
- Prior art keywords
- gas turbine
- turbine
- fuel
- model
- exhaust gas
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 68
- 230000008859 change Effects 0.000 claims abstract description 20
- 239000007789 gas Substances 0.000 claims description 152
- 239000000446 fuel Substances 0.000 claims description 80
- 230000006870 function Effects 0.000 claims description 27
- 238000004088 simulation Methods 0.000 claims description 26
- 238000002485 combustion reaction Methods 0.000 claims description 25
- 238000012546 transfer Methods 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 12
- 239000002918 waste heat Substances 0.000 claims description 12
- 230000014509 gene expression Effects 0.000 claims description 6
- 238000012795 verification Methods 0.000 claims description 6
- 239000012530 fluid Substances 0.000 claims description 3
- 239000007788 liquid Substances 0.000 claims description 3
- 238000010801 machine learning Methods 0.000 claims description 3
- 238000012549 training Methods 0.000 claims description 3
- 230000001419 dependent effect Effects 0.000 claims description 2
- 238000010977 unit operation Methods 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 8
- 239000011159 matrix material Substances 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 230000001172 regenerating effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/17—Mechanical parametric or variational design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/28—Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Analysis (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Fluid Mechanics (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Algebra (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Computational Mathematics (AREA)
- Feedback Control In General (AREA)
Abstract
The invention discloses a method for identifying key operation parameter online intervals of a gas turbine, which utilizes a gas turbine modeling method based on a Rowen model, and on the basis, provides that a Bayesian regression method is applied to the identification of key operation parameters of the gas turbine, and the key operation parameters can be identified in real time by carrying out parameter online interval identification on the real-time change of the working condition in the unit operation, so that the method is more suitable for the field working condition of the unit operation, and the identified parameters have the advantages of reasonability, accuracy and the like due to the unique interval identification characteristics of Bayesian regression.
Description
Technical Field
The invention belongs to the field of energy systems, and particularly relates to a gas turbine parameter identification method.
Background
The distributed energy system based on the gas turbine can effectively improve the energy utilization rate, reduce the environmental pollution emission, improve the power grid operation quality and effectively improve the energy utilization rate, thereby optimizing and adjusting the energy structure. The distributed energy system is a reasonable energy use mode, and has the important functions of saving energy, reducing emission, relieving power shortage, reducing peak-valley difference of gas and power grids, and improving power supply safety. Compared with the traditional centralized energy supply, the distributed energy system has the advantages that the form of energy supply is more dispersed, the energy supply can be determined according to the actual needs of local users, the utilization of various energy sources is integrated, and the cascade utilization of different energy sources and the comprehensive recovery and utilization of renewable energy sources are realized.
With the aid of computer technology, building gas turbine system models and performing digital simulations are becoming the most effective means for studying their dynamic characteristics. However, existing gas turbine modeling methods do not provide a method for model parameter estimation. Influence variables in the modeling of the gas turbine are complicated, and the problems of high identification limitation and low identification rate of structural parameters in nonlinear system identification are more prominent in practical engineering. How to effectively handle numerous and complicated uncertainty to obtain regularity cognition is a big problem, a gas turbine set model is a nonlinear model, and a parameter identification strategy is required to be adopted for the uncertainty of the operation of the gas turbine set model.
Disclosure of Invention
In order to solve the technical problems mentioned in the background art, the invention provides an online interval identification method for key operation parameters of a gas turbine.
In order to achieve the technical purpose, the technical scheme of the invention is as follows:
a method for identifying key operation parameters of a gas turbine on-line interval comprises the following steps:
(1) modeling a gas turbine by adopting a simulink/matlab platform based on a Rowen model, establishing two output systems including input and output power of fuel quantity and turbine exhaust gas temperature, and performing simulation verification;
(2) setting random interference of operation parameters, simulating field operation conditions, and acquiring output data of turbine exhaust gas temperature;
(3) adding interference of input fuel quantity into the established model of the gas turbine, carrying out a simulation experiment, and collecting sample data, wherein the sample data comprises input data and output data, the input data is the fuel quantity flow rate, and the output data is the turbine exhaust gas temperature;
(4) setting real-time dynamic working conditions, and performing online interval identification on parameters of a fuel quantity-turbine exhaust gas temperature transfer function model by adopting a Bayes regression method;
(5) and estimating the turbine exhaust gas temperature after the identification result of the transfer function parameters, the fuel quantity input change and the parameter random change by using a Bayesian regression method, and verifying the rationality of parameter identification according to the comparison of the operation parameters, the estimation result of the turbine exhaust gas temperature and the simulation data.
Further, in the step (1), based on a Rowen model, a fuel quantity input pipeline, a gas compressor, a combustion chamber, a waste heat delay module and a turbine volume module are established, thermodynamic characteristics and a control system of the gas turbine are integrated together, two characteristic algebraic equations of a control loop and the gas turbine are included, the establishment of each module adopts a transfer function expression method, each link is represented by different block units and then connected in series to form an integral model of the gas turbine.
Further, the specific process of step (1) is as follows:
(1a) the model establishment of the fuel quantity input pipeline is established based on two dynamic links of valve adjustment and pipeline volume, and the two dynamic links are simulated by a first-order inertia link:
wherein, KpAnd T0Is an inertia link parameter, and s is a pull operator;
for the adjustment of the liquid fuel valve, the fuel quantity entering the combustion chamber is adjusted by changing the bypass fuel, namely, the bypass adjusting valve is adjusted;
(1b) the compressibility of the fuel and the air generates an inertia time constant, and the process of releasing the gas by the gas compressor is simulated by a first-order inertia link;
(1c) the establishment of a combustion chamber and a waste heat delay module, wherein the delay exists between the fuel is injected into the combustion chamber and the combustion is started, and the combustion chamber is simulated by a delay link; there is a delay between the fuel combustion heat release to the thermocouple sensing the temperature change, called the waste heat delay, which is dependent on the size of the gas turbine and the average flow rate of the fluid;
(1d) the volume characteristic of a turbine and the establishment of a turbine volume module are realized, the turbine volume of the gas turbine is simulated by a first-order inertia link, the gas quantity change delay time parameter is calculated, and then the turbine volume module is established;
(1e) the output mechanical torque function and the exhaust temperature function are mainly established on the basis of two characteristic algebraic equations, the exhaust temperature, the mechanical torque, the fuel quantity and the rotor speed are linearly related, the exhaust temperature, the mechanical torque, the fuel quantity and the rotor speed are regarded as empirical functions of the fuel flow and the rotor speed, and the expressions are respectively as follows:
f1＝Tr-650(1-Wf1)+550(1-ω)
f2＝1.3(Wf2-0.23)+0.5(1-ω)
wherein, Wf1The amount of fuel at which the gas turbine reaches steady state at full load, f1Is Wf1Corresponding exhaust gas temperature; wf2The amount of fuel at which the gas turbine reaches steady state during idling, f2Is Wf2A corresponding mechanical torque; t isrIs an exhaust temperature reference value; omega is the rotor speed;
(1f) and connecting the built modules according to the operation logic of the gas turbine, adding the change of the flow rate of the fuel, performing simulation verification, and analyzing the output of the turbine exhaust gas temperature.
Further, the delay of the combustion chamber and the waste heat delay are both in the order of ms, and the combustion chamber and the waste heat delay module are omitted when the model is built.
Further, the specific process of step (2) is as follows:
(2a) adding fuel quantity and flow velocity random interference, simulating the loading and unloading conditions during field operation, and acquiring output data of turbine exhaust gas temperature;
(2b) adding an inertia link parameter KpThe random interference of the turbine is realized, the real situation of field operation is simulated, and the output data of the turbine exhaust gas temperature is obtained;
(2c) adding inertia link parameter T0The random interference of the turbine is used for simulating the real situation of field operation and obtaining the output data of the turbine exhaust gas temperature.
Further, in the step (3), based on a simulink platform, a simulation experiment is performed according to the established model, interference of input fuel quantity is added, an output curve of turbine exhaust gas temperature is observed, the output curve is found to accord with the output characteristic of a first-order inertia link, and then sample data including fuel quantity flow rate and turbine exhaust gas temperature are collected.
Further, the specific process of step (4) is as follows:
(4a) selecting a first-order inertia link model as a simplified model of the gas turbine by adopting a Bayesian regression method in machine learning, and identifying input and output parameters and relations;
(4b) taking the fuel quantity as an input quantity, taking the turbine exhaust gas temperature as an output quantity, taking data in the simulation model as a training set, adopting a first-order inertia link model system, carrying out parameter identification by using a Bayesian regression algorithm, and identifying a real-time interval of key operation parameters of the dynamic system model;
(4c) taking the input flow rate of the fuel quantity in the gas turbine as a variable working condition, thereby simulating the real-time variable input and output characteristics of the gas turbine, and identifying the key operation parameters of the gas turbine in an online state on the basis;
(4d) the random variation of the key parameters in the gas turbine is used as a variable working condition, so that the real-time variation input and output characteristics of the gas turbine are simulated, on the basis, the key operation parameters of the gas turbine in an online state are identified, and an online interval self-identification result of the key operation parameters of the gas turbine is obtained.
Further, the specific process of step (5) is as follows:
(5a) performing online interval self-identification on key parameters of the gas turbine by using a Bayesian regression method to obtain an identification result of the operation parameters and interval estimation of the exhaust gas temperature of the gas turbine according to the identification result of the operation parameters, and comparing and analyzing the reasonability and accuracy of the parameter identification result;
(5b) according to the parameter K of the inertia link after random changepComparing the original data with the identification result of the Bayesian regression method, and analyzing the rationality of the identification result;
(5c) according to the parameter T of the inertia link after random change0And comparing the original data with the identification result of the Bayesian regression method, and analyzing the rationality of the identification result.
Adopt the beneficial effect that above-mentioned technical scheme brought:
the invention utilizes a Rowen model-based gas turbine modeling method, and provides a Bayesian regression method applied to the identification of key operation parameters of the gas turbine on the basis, and the key operation parameters can be identified in real time by carrying out parameter online interval identification on the real-time change of the operating conditions of the unit during operation, so that the method is more suitable for the field operating conditions of the unit operation, and the identified parameters have the advantages of more reasonability and accuracy and the like due to the unique interval identification characteristics of Bayesian regression.
Drawings
FIG. 1 is a flow chart of the principle of online identification of a Bayesian regression method of the present invention;
FIG. 2 is a basic flow diagram of the present invention;
FIG. 3 is a schematic model diagram of a simulink platform of a Rowen model based gas turbine cycle system in an embodiment;
FIG. 4 is a graph of a random step change in fuel quantity (flow rate) in an example;
FIG. 5 is a diagram illustrating the fitting effect of a Bayesian regression method on estimated and true turbine exhaust temperature values after online self-identification of operating parameters in an embodiment;
FIG. 6 is a diagram illustrating the Bayesian regression method on the model parameter K of the inertial link in the embodimentpOn-line self-identifying estimated and true valuesA schematic diagram of the fitting effect;
FIG. 7 is a diagram illustrating the Bayesian regression method applied to the parameter T in the inertia link model in the embodiment0And (5) on-line self-identification of the fitting effect schematic diagram of the estimated value and the true value.
Detailed Description
The technical scheme of the invention is explained in detail in the following with the accompanying drawings.
The invention designs an online interval identification method for key operating parameters of a gas turbine, which comprises the following steps as shown in figures 1-2:
s1: modeling a gas turbine by adopting a simulink/matlab platform based on a Rowen model, establishing two output systems including input and output power of fuel quantity and turbine exhaust gas temperature, and performing simulation verification;
s2: setting random interference of operation parameters, simulating field operation conditions, and acquiring output data of turbine exhaust gas temperature;
s3: adding interference of input fuel quantity into the established model of the gas turbine, carrying out a simulation experiment, and collecting sample data, wherein the sample data comprises input data and output data, the input data is the fuel quantity flow rate, and the output data is the turbine exhaust gas temperature;
s4: setting real-time dynamic working conditions, and performing online interval identification on parameters of a fuel quantity-turbine exhaust gas temperature transfer function model by adopting a Bayes regression method;
s5: and estimating the turbine exhaust gas temperature after the identification result of the transfer function parameters, the fuel quantity input change and the parameter random change by using a Bayesian regression method, and verifying the rationality of parameter identification according to the comparison of the operation parameters, the estimation result of the turbine exhaust gas temperature and the simulation data.
As a further preferred embodiment, the step S1 specifically includes:
the modeling of the gas turbine is based on a Rowen model, a fuel quantity input pipeline, a gas compressor, a combustion chamber, a turbine, a volume module and the like are mainly established, the thermodynamic characteristics and a control system of the gas turbine are integrated together, and the gas turbine comprises a control loop and two characteristic algebraic equations of the gas turbine. The establishment of each module gets rid of the traditional method of expressing the dynamic characteristic of the gas turbine by using a differential equation set, mainly adopts an expression method of a transfer function, simultaneously refers to the modularized modeling idea, expresses a system mathematical model by using a block diagram, expresses each link by using different block units, and then forms an integral model of the gas turbine by connecting in series, as shown in fig. 3.
The application range of the Rowen model is as follows: (1) the ambient temperature is 15 ℃, and the atmospheric pressure is 101.35 kPa; (2) the relative speed of the rotor is within the range of 95-107% of the rated rotating speed; (3) a simple cycle non-regenerative gas turbine; (4) the method does not consider the starting and stopping processes of the gas turbine and is used for researching the dynamic characteristics of the gas turbine after the gas turbine enters the steady-state operation.
S11: the model establishment of the fuel quantity input pipeline is established based on two dynamic links of valve adjustment and pipeline volume, wherein the two dynamic links are simulated by a first-order inertia link,
for the adjustment of the liquid fuel valve, the fuel quantity entering the combustion chamber is adjusted by changing the bypass fuel, namely, the bypass adjusting valve is adjusted; the values of the model parameters can be set by referring to data provided by manufacturers, and are specifically determined for corresponding numerical values in the model;
s12: the compressibility of the fuel and the air can generate an inertia time constant, and the process of releasing the gas by the gas compressor is simulated by a first-order inertia link;
s13: the establishment of a combustion chamber and waste heat delay module, there is a delay between the fuel being injected into the combustion chamber and the start of combustion, so the combustion chamber is modeled as a delay chain, which in modern gas turbines is of the order of ms; there is a time delay between the fuel combustion heat release to the thermocouple sensing the temperature change, called the waste heat delay, which depends on the size of the gas turbine and the average flow rate of the fluid, and is generally about ms magnitude, because these two delay times are short and have little influence on the model simulation, these two delay modules are omitted in the model;
s14: regarding the volume characteristic of a turbine and the establishment of a turbine volume module, the turbine volume of a gas turbine is a non-negligible part, according to various reference data of a fuel system, a first-order inertia link is used for simulation, a gas quantity change delay time parameter is calculated, and then the turbine volume module is established;
s15: the output mechanical torque function and the exhaust temperature function are mainly established on the basis of two characteristic algebraic equations, the exhaust temperature, the mechanical torque, the fuel flow and the rotor rotating speed are linearly related, the exhaust temperature, the mechanical torque, the fuel flow and the rotor rotating speed are regarded as empirical functions of the fuel flow and the rotating speed, and the expressions are respectively as follows:
f1＝Tr-650(1-Wf1)+550(1-ω)
f2＝1.3(Wf2-0.23)+0.5(1-ω)
when the full load of the gas turbine reaches the steady state, the fuel quantity Wf1Is 1 (per unit value), the rotor speed reaches the rated speed, and the turbine exhaust temperature is the exhaust temperature reference value Tr(exhaust temperature reference 950 ℃ in this model); when the gas turbine reaches steady state at no load, the fuel quantity Wf20.23 (per unit), the rotor speed reaches the rated speed 1 (per unit), and the gas turbine output mechanical torque is 0.
S16: and connecting the built modules according to the operation logic of the gas turbine, adding the change of input quantity (fuel quantity and flow rate), performing simulation verification, and analyzing the output of the turbine exhaust gas temperature, so that the general rule of the operation of the gas turbine is met.
As a further preferred embodiment, the step S2 specifically includes:
s21: adding random interference of model input data (fuel quantity and flow velocity), simulating the loading and load shedding conditions during field operation, and acquiring output data of the exhaust gas temperature;
s22: adding an inertia link parameter KpThe random interference of the system simulates the real situation of field operation and acquires the output data of the exhaust gas temperature;
s23: adding inertia link parameter T0The random interference of the system simulates the real situation of field operation and acquires the output data of the exhaust gas temperature.
As a further preferred embodiment, the step S3 specifically includes:
the simulation experiment is based on a simulink platform, the simulation experiment is carried out according to the established model, the interference of the input fuel quantity is added, and the operation parameter K of the inertia link is addedp、T0The random interference of (2) ensures the coverage of sample points, and the working conditions of the simulation operation mainly comprise:
s41: the fuel quantity of the gas turbine is stepped from 0kg/s to 1.72kg/s at 0 s;
s42: the fuel quantity of the gas turbine is stepped from 1.72kg/s to 4.3kg/s at 10 s;
s43: the fuel quantity of the gas turbine is stepped from 4.3kg/s to 2.58kg/s at 20 s;
s44: the fuel quantity of the gas turbine is stepped from 2.58kg/s to 3.44kg/s at 30 s;
s45: the fuel quantity of the gas turbine was stepped from 3.44kg/s to 2.15kg/s at 40 s.
The operation condition of the gas turbine under the real-time working condition is simulated, the collected sample data comprises input data (fuel quantity flow rate) and output data (turbine exhaust gas temperature), and the sampling time of the measured data is set to be 0.1 second, as shown in fig. 4.
As a further preferred embodiment, the step S4 specifically includes:
s41: and selecting a proper function model by adopting a Bayesian estimation algorithm in machine learning, such as: as a simplified model of the gas turbine, a standard transfer function model of a self-balanced inertial system is used, and in this embodiment, a first-order inertial transfer function model is used to identify input and output parameters and relationships:
s42: according to the Bayes principle, the method for the detection of the biological characteristic,
by prior probability
Determining a posterior probability, using data of the training number set as a likelihood function
The normalization constant p (D) is a constant;
s43: for the system model, first for the first order integral inertia element:
discretizing system simulation data:
wherein, U (t) is used as the input value of the system, Y (t-1) is the system output data value at the time of t-1, and Y (t) is the system output data value at the time of t;
s44: will be provided with
Viewed as ky1，
Viewed as ky2Then k for one yy1x1+ky2x2The type model is more convenient for parameter identification of Bayesian estimation;
s45: the method comprises the following steps of utilizing a Bayesian regression method to carry out online interval self-identification on key operation parameters of the gas turbine, wherein the Bayesian regression method comprises the following specific steps:
s451: first, a distribution of conjugate priors is constructed (i.e., a distribution assuming an initial prior)
) When the likelihood function of the observed data is multiplied by the prior, the function with the same form as the prior is obtained, and the function is obtainedA simple update equation of the conjugate prior parameter converts it to a posterior distribution. This enables the estimates to be updated in order as new data arrives. Only need to calculate according to formula
And
updating of (1);
S452：
and
the updating of the parameters is performed by the following formula (for an nth order matrix):
wherein the content of the first and second substances,
is an input data matrix (i.e., flow rate of fuel quantity);
s453: then the posterior of the previous step is updated for the prior of the next step
And
of the last step
And
is the next step
And
by adopting a recursion mode, the loop updating can be carried out through Bayesian estimation,
as a variance matrix of the bayesian estimate and the true value,
namely, the parameter matrix to be identified can obtain the interval identification result of the parameter to be identified by utilizing the normalized form given by the following formula:
s46: by passing
The accuracy of Bayesian estimation is evaluated as an index,
is the k soughtyiBy back-deriving the model parameters T defined in this document0、KpNamely:
s47: and substituting the simulation data value of the fuel quantity-output power, and compiling and performing a program of a Bayesian regression method through matlab, so that the online interval identification of the operation parameters can be realized.
As a further preferred embodiment, in step S6, specifically, the method includes:
s61: the recognition result of the first-order transfer function parameters is estimated by the Bayes regression method, the exhaust gas temperature after the fuel quantity input change and the parameter random change is estimated, and the exhaust gas temperature of the recognition result is compared with the simulation result, as shown in FIG. 5, the recognition result is found to be accurate, the error of the output curve is small, and the Bayes regression recognition is proved to have rationality and accuracy.
S62: the results of the first order transfer function parameter identification using the bayesian regression method, the gas turbine operating parameter estimation results and the simulation data are compared, as shown in figures 6-7, the identification result is more accurate, the confidence interval of the identification result basically comprises real parameter values, the Bayesian regression identification is proved to have rapidity and accuracy, the Bayesian regression method is adopted to identify the parameters as real-time identification, can accurately identify the random variation of the operation parameters, can identify the random variation interval of the key operation parameters of the gas turbine in real time, lays a good foundation for establishing a control model of the gas turbine, and analyzes the uncertainty of the identification result of the operation parameters of the gas turbine, and a novel method is provided for identifying and predicting key parameters of the gas turbine such as future load, future exhaust gas temperature, output power and the like.
The embodiments are only for illustrating the technical idea of the present invention, and the technical idea of the present invention is not limited thereto, and any modifications made on the basis of the technical scheme according to the technical idea of the present invention fall within the scope of the present invention.
Claims (8)
1. A method for identifying key operation parameters of a gas turbine on-line interval is characterized by comprising the following steps:
(1) modeling a gas turbine by adopting a simulink/matlab platform based on a Rowen model, establishing two output systems including input and output power of fuel quantity and turbine exhaust gas temperature, and performing simulation verification;
(2) setting random interference of operation parameters, simulating field operation conditions, and acquiring output data of turbine exhaust gas temperature;
(3) adding interference of input fuel quantity into the established model of the gas turbine, carrying out a simulation experiment, and collecting sample data, wherein the sample data comprises input data and output data, the input data is the fuel quantity flow rate, and the output data is the turbine exhaust gas temperature;
(4) setting real-time dynamic working conditions, and performing online interval identification on parameters of a fuel quantity-turbine exhaust gas temperature transfer function model by adopting a Bayes regression method;
(5) and estimating the turbine exhaust gas temperature after the identification result of the transfer function parameters, the fuel quantity input change and the parameter random change by using a Bayesian regression method, and verifying the rationality of parameter identification according to the comparison of the operation parameters, the estimation result of the turbine exhaust gas temperature and the simulation data.
2. The method for identifying the key operating parameters of the gas turbine as claimed in claim 1, wherein in the step (1), the fuel input pipeline, the compressor, the combustion chamber, the waste heat delay module and the turbine volume module are built based on a Rowen model, thermodynamic characteristics and a control system of the gas turbine are integrated together, two characteristic algebraic equations of a control loop and the gas turbine are included, the building of each module adopts a transfer function expression method, each link is represented by different block units and then is connected in series to form an integral model of the gas turbine.
3. The method for identifying the key operation parameters of the gas turbine on-line interval as claimed in claim 2, wherein the specific process of the step (1) is as follows:
(1a) the model establishment of the fuel quantity input pipeline is established based on two dynamic links of valve adjustment and pipeline volume, and the two dynamic links are simulated by a first-order inertia link:
wherein，KpAnd T0Is an inertia link parameter, and s is a pull operator;
for the adjustment of the liquid fuel valve, the fuel quantity entering the combustion chamber is adjusted by changing the bypass fuel, namely, the bypass adjusting valve is adjusted;
(1b) the compressibility of the fuel and the air generates an inertia time constant, and the process of releasing the gas by the gas compressor is simulated by a first-order inertia link;
(1c) the establishment of a combustion chamber and a waste heat delay module, wherein the delay exists between the fuel is injected into the combustion chamber and the combustion is started, and the combustion chamber is simulated by a delay link; there is a delay between the fuel combustion heat release to the thermocouple sensing the temperature change, called the waste heat delay, which is dependent on the size of the gas turbine and the average flow rate of the fluid;
(1d) the volume characteristic of a turbine and the establishment of a turbine volume module are realized, the turbine volume of the gas turbine is simulated by a first-order inertia link, the gas quantity change delay time parameter is calculated, and then the turbine volume module is established;
(1e) the output mechanical torque function and the exhaust temperature function are mainly established on the basis of two characteristic algebraic equations, the exhaust temperature, the mechanical torque, the fuel quantity and the rotor speed are linearly related, the exhaust temperature, the mechanical torque, the fuel quantity and the rotor speed are regarded as empirical functions of the fuel flow and the rotor speed, and the expressions are respectively as follows:
f1＝Tr-650(1-Wf1)+550(1-ω)
f2＝1.3(Wf2-0.23)+0.5(1-ω)
wherein, Wf1The amount of fuel at which the gas turbine reaches steady state at full load, f1Is Wf1Corresponding exhaust gas temperature; wf2The amount of fuel at which the gas turbine reaches steady state during idling, f2Is Wf2A corresponding mechanical torque; t isrIs an exhaust temperature reference value; omega is the rotor speed;
(1f) and connecting the built modules according to the operation logic of the gas turbine, adding the change of the flow rate of the fuel, performing simulation verification, and analyzing the output of the turbine exhaust gas temperature.
4. The method for identifying the key operation parameters of the gas turbine engine on-line interval as claimed in claim 3, wherein the delay of the combustion chamber and the delay of the waste heat are both in ms magnitude, and the delay modules of the combustion chamber and the waste heat are omitted when the model is constructed.
5. The method for identifying the key operation parameters of the gas turbine on-line interval as claimed in claim 3, wherein the specific process of the step (2) is as follows:
(2a) adding fuel quantity and flow velocity random interference, simulating the loading and unloading conditions during field operation, and acquiring output data of turbine exhaust gas temperature;
(2b) adding an inertia link parameter KpThe random interference of the turbine is realized, the real situation of field operation is simulated, and the output data of the turbine exhaust gas temperature is obtained;
(2c) adding inertia link parameter T0The random interference of the turbine is used for simulating the real situation of field operation and obtaining the output data of the turbine exhaust gas temperature.
6. The method for identifying the key operation parameters of the gas turbine on-line interval according to the claim 3 is characterized in that in the step (3), based on a simulink platform, a simulation experiment is carried out according to an established model, interference of input fuel quantity is added, an output curve of turbine exhaust gas temperature is observed, the output characteristic of the turbine exhaust gas temperature is found to be in accordance with a first-order inertia link, and then sample data including fuel quantity flow rate and turbine exhaust gas temperature are collected.
7. The method for identifying the key operation parameters of the gas turbine on-line interval as claimed in claim 3, wherein the specific process of the step (4) is as follows:
(4a) selecting a first-order inertia link model as a simplified model of the gas turbine by adopting a Bayesian regression method in machine learning, and identifying input and output parameters and relations;
(4b) taking the fuel quantity as an input quantity, taking the turbine exhaust gas temperature as an output quantity, taking data in the simulation model as a training set, adopting a first-order inertia link model system, carrying out parameter identification by using a Bayesian regression algorithm, and identifying a real-time interval of key operation parameters of the dynamic system model;
(4c) taking the input flow rate of the fuel quantity in the gas turbine as a variable working condition, thereby simulating the real-time variable input and output characteristics of the gas turbine, and identifying the key operation parameters of the gas turbine in an online state on the basis;
(4d) the random variation of the key parameters in the gas turbine is used as a variable working condition, so that the real-time variation input and output characteristics of the gas turbine are simulated, on the basis, the key operation parameters of the gas turbine in an online state are identified, and an online interval self-identification result of the key operation parameters of the gas turbine is obtained.
8. The method for identifying the key operation parameters of the gas turbine on-line interval as claimed in claim 3, wherein the specific process of the step (5) is as follows:
(5a) performing online interval self-identification on key parameters of the gas turbine by using a Bayesian regression method to obtain an identification result of the operation parameters and interval estimation of the exhaust gas temperature of the gas turbine according to the identification result of the operation parameters, and comparing and analyzing the reasonability and accuracy of the parameter identification result;
(5b) according to the parameter K of the inertia link after random changepComparing the original data with the identification result of the Bayesian regression method, and analyzing the rationality of the identification result;
(5c) according to the parameter T of the inertia link after random change0And comparing the original data with the identification result of the Bayesian regression method, and analyzing the rationality of the identification result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110372207.1A CN113111456B (en) | 2021-04-07 | Online interval identification method for key operation parameters of gas turbine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110372207.1A CN113111456B (en) | 2021-04-07 | Online interval identification method for key operation parameters of gas turbine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113111456A true CN113111456A (en) | 2021-07-13 |
| CN113111456B CN113111456B (en) | 2024-07-23 |
Family
ID=
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113885311A (en) * | 2021-09-18 | 2022-01-04 | 东南大学溧阳研究院 | Closed-loop identification method for combustion chamber model of gas turbine based on generalized frequency method |
| CN116127716A (en) * | 2022-12-19 | 2023-05-16 | 华北电力科学研究院有限责任公司 | Method and device for identifying flow characteristics of valve of steam turbine |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101078373A (en) * | 2007-07-05 | 2007-11-28 | 东北大学 | Combustion controlling device and controlling method for mini combustion turbine |
| CN101328836A (en) * | 2008-07-04 | 2008-12-24 | 东南大学 | Multi-model self-adapting generalized forecast control method of gas turbine rotary speed system |
| JP2009068359A (en) * | 2007-09-11 | 2009-04-02 | Japan Aerospace Exploration Agency | Performance estimation system of gas turbine engine |
| US20110224959A1 (en) * | 2008-10-17 | 2011-09-15 | Zhang Xiao-Yi | Gas turbine model and a method for the modeling of a gas turbine |
| CN102996424A (en) * | 2011-09-15 | 2013-03-27 | 通用电气公司 | System and method for simulating a gas turbine compressor |
| CN106682376A (en) * | 2017-04-01 | 2017-05-17 | 国网河南省电力公司电力科学研究院 | Whole-process steam turbine modeling and recognizing method of actual characteristics of parameters changing with working conditions |
| CN109143892A (en) * | 2018-09-06 | 2019-01-04 | 国网山西省电力公司电力科学研究院 | A method of considering prime mover and its Speed-adjustable system parameter identification of idle influence |
| CN110826021A (en) * | 2019-10-31 | 2020-02-21 | 哈尔滨工业大学 | Robust identification and output estimation method for nonlinear industrial process |
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101078373A (en) * | 2007-07-05 | 2007-11-28 | 东北大学 | Combustion controlling device and controlling method for mini combustion turbine |
| JP2009068359A (en) * | 2007-09-11 | 2009-04-02 | Japan Aerospace Exploration Agency | Performance estimation system of gas turbine engine |
| CN101328836A (en) * | 2008-07-04 | 2008-12-24 | 东南大学 | Multi-model self-adapting generalized forecast control method of gas turbine rotary speed system |
| US20110224959A1 (en) * | 2008-10-17 | 2011-09-15 | Zhang Xiao-Yi | Gas turbine model and a method for the modeling of a gas turbine |
| CN102996424A (en) * | 2011-09-15 | 2013-03-27 | 通用电气公司 | System and method for simulating a gas turbine compressor |
| CN106682376A (en) * | 2017-04-01 | 2017-05-17 | 国网河南省电力公司电力科学研究院 | Whole-process steam turbine modeling and recognizing method of actual characteristics of parameters changing with working conditions |
| CN109143892A (en) * | 2018-09-06 | 2019-01-04 | 国网山西省电力公司电力科学研究院 | A method of considering prime mover and its Speed-adjustable system parameter identification of idle influence |
| CN110826021A (en) * | 2019-10-31 | 2020-02-21 | 哈尔滨工业大学 | Robust identification and output estimation method for nonlinear industrial process |
Non-Patent Citations (1)
| Title |
|---|
| 黄玉柱 等: "一种重型燃气轮机建模及其参数估计的方法", 《动力工程学报》, vol. 36, no. 8 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113885311A (en) * | 2021-09-18 | 2022-01-04 | 东南大学溧阳研究院 | Closed-loop identification method for combustion chamber model of gas turbine based on generalized frequency method |
| CN113885311B (en) * | 2021-09-18 | 2024-04-16 | 东南大学溧阳研究院 | Gas turbine combustion chamber model closed-loop identification method based on generalized frequency method |
| CN116127716A (en) * | 2022-12-19 | 2023-05-16 | 华北电力科学研究院有限责任公司 | Method and device for identifying flow characteristics of valve of steam turbine |
| CN116127716B (en) * | 2022-12-19 | 2024-03-26 | 华北电力科学研究院有限责任公司 | Method and device for identifying flow characteristics of valve of steam turbine |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7593804B2 (en) | Fixed-point virtual sensor control system and method | |
| US20100325071A1 (en) | System and method for empirical ensemble-based virtual sensing of gas emission | |
| Duan et al. | Nonlinear modeling of regenerative cycle micro gas turbine | |
| CN104090491B (en) | Gas steam combined cycle unit multivariable constrained prediction function load control method | |
| CN103559537B (en) | Based on the template matching method of error back propagation in a kind of out of order data stream | |
| CN104539601B (en) | Dynamic network attack process analysis method for reliability and system | |
| CN112380630A (en) | New energy automobile battery thermal runaway risk prediction method based on Internet of vehicles data | |
| Pal et al. | Constrained surrogate-based engine calibration using lower confidence bound | |
| CN113361214A (en) | Open channel control model parameter identification method based on water level flow data | |
| CN112016754A (en) | Power station boiler exhaust gas temperature advanced prediction system and method based on neural network | |
| EP2562611A2 (en) | Process for adaptive modeling of performance degradation | |
| Zeng et al. | A study on extreme learning machine for gasoline engine torque prediction | |
| CN113111456B (en) | Online interval identification method for key operation parameters of gas turbine | |
| CN113111456A (en) | Online interval identification method for key operating parameters of gas turbine | |
| CN109783882A (en) | A kind of gas turbine fuel system modeling and simulating method for combining matlab and flowmaster | |
| CN111275320B (en) | Performance adjustment data processing method, system and storage medium of generator set | |
| JP6032847B2 (en) | Steam turbine start control system | |
| CN112182905A (en) | Heat supply pipe network simulation method and device for comprehensive energy system | |
| CN110684547A (en) | Optimized control method for biomass pyrolysis carbonization kiln | |
| CN115291509A (en) | Gas turbine dynamic parameter online identification method based on hybrid deep learning model | |
| CN100366876C (en) | Online analysis method and system for operation efficiency of combined gas-steam cycle power station | |
| CN112670997B (en) | Electric heating energy system time sequence probability power flow calculation method considering photovoltaic uncertainty | |
| CN114810337A (en) | Control method and related device of supercharger | |
| Liu et al. | Time series prediction with input noise based on the ESN and the EM and its industrial applications | |
| CN117912585B (en) | Optimization method for combustion chemical reaction based on deep artificial neural network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Sun Li Inventor after: Sun Shoutai Inventor after: Li Shi Inventor after: Xue Yali Inventor before: Sun Shoutai Inventor before: Li Shi Inventor before: Sun Li Inventor before: Xue Yali |
|
| GR01 | Patent grant |