CN105912878A - Gas turbine adaptive gas circuit component performance diagnostic method based on combination of thermal model and particle swarm optimization - Google Patents

Abstract

The purpose of the present invention is to provide a gas turbine adaptive gas path component performance diagnosis method based on the combination of thermal model and particle swarm optimization algorithm, establish a non-linear thermal model of gas turbine, and redefine the gas path health index of compressor and turbine with similar conversion parameters , collect the gas path measurement parameters of a certain period of time when the current target gas turbine is running stably, and use it as the gas path measurement parameters to be diagnosed offline after noise reduction processing, and obtain the current gas path of each component through the iterative optimization calculation of the particle swarm optimization algorithm The health index is used to evaluate the actual performance health status of the target gas turbine. The invention solves the problem that the diagnostic accuracy of the traditional gas turbine gas circuit component performance diagnosis method is easily affected by the environmental conditions and operating conditions, improves the local optimization characteristics of the traditional diagnostic algorithm, improves the accuracy of the diagnostic results, and simplifies the diagnostic process. , which can be effectively applied to the situation where there is measurement noise and the performance diagnosis of complex gas turbine units.

Description

Gas turbine adaptive gas path component performance diagnosis method based on thermal model and particle swarm optimization algorithm Law

technical field

The invention relates to a diagnostic method, in particular to a gas turbine performance diagnostic method.

Background technique

In the second half of the 20th century, with the widespread application of gas turbines in the aviation industry, more and more attention has been paid to the field of industrial power stations, oil and gas pipeline transportation, and the shipbuilding industry. During operation, due to the harsh working conditions of high temperature, high pressure, high speed and high stress, and environmental pollution, the performance of various types of gas turbines will gradually decline. The main gas circuit components of a gas turbine include a compressor, a combustor and a turbine. These major components are subject to different degradation phenomena over time, such as dirt, leakage, corrosion, thermal distortion, foreign object damage, etc., which will cause performance deterioration and easily lead to various serious failures and operational safety issues. Therefore, the current gas turbine performance health is very important information for gas turbine users.

At present, the maintenance strategy of most gas turbines is preventive maintenance, which usually considers the need for minor repairs, medium repairs, and major repairs according to the equivalent operating hours (EOH) indicated by the gas turbine manufacturer. Gas turbine outages, whether planned or unplanned, are always costly. In order to save maintenance costs, users need to adopt maintenance strategies based on the actual performance and health of gas turbines, that is, predictive maintenance.

The performance diagnosis method of gas circuit components based on thermal model decision has been widely used in gas turbine performance and health monitoring, and has become one of the key technologies to support the reform of maintenance strategy. The usual gas path component performance diagnosis method based on thermal model decision-making uses the performance parameters (absolute parameters) of gas path components to define the component health index, so the gas path measurement parameters need to be preprocessed to eliminate environmental conditions or operating conditions before diagnosis. And the influence of the change of the composition of the working fluid on the performance of the gas turbine. And because the component performance parameters are used as adaptive variables, the diagnosis usually needs to be divided into two steps: the first step is to calculate the component performance parameters through linear or nonlinear Newton-Lepson iterative algorithm according to the gas path measurement parameters, such as air flow rate , compressor pressure ratio, compressor isentropic efficiency, turbine front temperature, turbine isentropic efficiency and other absolute performance parameters; the second step is to compare the operating point of the component under the actual performance degradation and the health reference condition on the same component characteristic map The operating point of the lower component, so as to observe the degree of deviation of the characteristic line on the component characteristic map at this time (that is, to obtain the health index of the gas path component), so as to evaluate the performance health of the current component. When the number of components participating in the diagnosis in the gas turbine increases, the dimension of the failure coefficient matrix will increase accordingly, coupled with the interference of measurement noise, the blurring effect (that is, although some components do not actually degrade in performance, the diagnosed The degradation in performance is distributed across almost all components (gas path health index) can be so strong that the components that actually degrade in performance are not identified.

Contents of the invention

The purpose of the present invention is to provide a gas turbine self-adaptive gas path component performance diagnosis method based on the combination of thermodynamic model and particle swarm optimization algorithm, which can be effectively applicable to measurement noise and complex gas turbine units.

The purpose of the present invention is achieved like this:

The present invention is based on the gas turbine self-adaptive gas path component performance diagnosis method based on the combination of thermal model and particle swarm optimization algorithm, which is characterized in that:

(1) Establish a nonlinear thermal model of the gas turbine based on the gas path measurement parameters when the target gas turbine is newly put into operation or healthy, in which the compressor and the turbine are expressed in the form of similar converted parameters;

(2) Redefine the gas path health index of compressors and turbines with similar conversion parameters to eliminate the impact of changes in gas turbine operating performance due to changes in environmental conditions;

(3) Collect gas path measurement parameters for a period of time when the current target gas turbine is running stably, and use it as gas path measurement parameters to be diagnosed offline after noise reduction processing;

(4) Set the environmental input conditions and operating input conditions of the established gas turbine thermal model to be consistent with the operating conditions of the target gas turbine during sampling, and eliminate the influence of changes in the operating performance of the gas turbine due to changes in environmental conditions and operating conditions;

(5) Taking the root mean square error between the gas path measurement parameters to be diagnosed offline and the gas path parameter data calculated by the thermal model as the objective function, the current gas path health of each component is obtained through iterative optimization calculation of the particle swarm optimization algorithm Index, used to evaluate the actual performance health of the target gas turbine.

The present invention may also include:

1. The specific steps for establishing the nonlinear thermal model of the gas turbine in step (1) are as follows:

(a) Establish a gas turbine component-level thermal model using the relative conversion parameters of the components, in which the compressor and turbine characteristic line data are organized into a common relative conversion parameter form:

The characteristic line of the compressor is sorted into a common similar converted parameter form as follows:

G _C,cor,rel ＝f(n _cor,rel ,π _C,rel )

η _C,rel = f(n _cor,rel ,π _C,rel )

in is the relative converted speed, n is the actual speed, is the inlet stagnation temperature of the compressor, R _g is the gas constant of the working fluid flowing through the compressor, and the subscript 0 indicates the design point;

is the relative equivalent flow rate, G _C is the actual compressor inlet flow rate, is the compressor inlet stagnation pressure, is the relative pressure ratio, π _C is the actual compressor pressure ratio, η _{C, rel} = η _C / η _C is the relative isentropic efficiency, and η _C is the actual compressor isentropic efficiency;

The turbine characteristic line is sorted into a common similar converted parameter form as follows:

G _T,cor,rel ＝f(n _cor,rel ,π _T,rel )

η _T,rel = f(n _cor,rel ,π _T,rel )

In the formula: is the relative converted speed, n is the actual speed, is the stagnation temperature at the inlet of the turbine, R _g is the gas constant of the working fluid flowing through the turbine, is the relative converted flow, G _T is the actual turbine inlet flow, is the relative pressure ratio, ηT _,rel = _ηT / _ηT0 is the relative isentropic efficiency, _ηT is the actual turbine isentropic efficiency, and the subscript 0 indicates the design point;

(b) According to the gas path measurement parameters collected when the target gas turbine is newly put into operation or healthy, gradually correct the characteristic line data of each component, so that the calculated value of the built gas turbine thermal model matches the measured parameters of the gas path, thereby eliminating the heat Negative impact of model calculation errors on diagnostic results.

2. Step (2) The specific steps to redefine the gas path health index of the compressor and turbine with similar conversion parameters are as follows:

(a) The overall performance and health status of the gas turbine is represented by the gas path health index of each component, and the gas path health index of the compressor and turbine is redefined with relative conversion parameters to eliminate the negative impact on the diagnosis results due to changes in environmental conditions. influences;

(b) The compressor air path health index is defined as follows:

SF _C,FC ＝G _{C,cor,rel,deg} /G _C,cor,rel

ΔSF _C,FC ＝(G _{C,cor,rel,deg} -G _C,cor,rel )/ _GC,cor,rel

SF _C,Eff =η _C,deg /η _C

ΔSF _C,Eff =(η _C,deg -η _C )/η _C

Among them, SF _{C, FC} is the flow characteristic index of the compressor; G _{C, cor, rel, deg} is the relative equivalent flow rate when the compressor performance declines; G _{C, cor, rel} is the relative equivalent flow rate when the compressor is healthy; SF _{C, Eff} is Compressor efficiency characteristic index; η _C,deg is the isentropic efficiency when the compressor performance declines; η _C is the isentropic efficiency when the compressor is healthy;

The combustion chamber gas path health index is defined as follows:

SF _B,Eff =η _B,deg /η _B

ΔSF _B,Eff =(η _B,deg -η _B )/η _B

Wherein SF _{B, Eff} is the combustion efficiency performance index of the combustion chamber; η _{B, deg} is the combustion efficiency when the performance of the combustion chamber declines; η _B is the combustion efficiency when the combustion chamber is healthy;

The turbine gas path health index is defined as follows:

SF _T,FC ＝G _T,cor,deg /G _T,cor

ΔSF _T,FC ＝(G _T,cor,deg -G _T,cor )/G _T,cor

SF _T,Eff =η _T,deg /η _T

ΔSF _T,Eff =(η _T,deg -η _T )/η _T

Among them, SF _T,FC is the turbine flow performance index; G _T,cor,deg is the equivalent flow rate when the turbine performance declines; G _T,cor is the equivalent flow rate when the turbine is healthy; SF _T,Eff is the turbine efficiency performance index; η _T,deg is the isentropic efficiency when the turbine performance declines; η _T is the isentropic efficiency when the turbine is healthy.

3. Step (5) takes the root mean square error between the gas path measurement parameters to be diagnosed offline and the gas path parameter data calculated by the thermal model as the objective function, and obtains the current parameters of each component through iterative optimization calculation of the particle swarm optimization algorithm. Gas path health index, the specific steps for evaluating the actual performance and health status of the target gas turbine are as follows:

(a) With gas path measurement data to be diagnosed offline Gas path parameter data calculated with thermal model The root mean square error between is the objective function Fitness:

$\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; no</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; =</span>\sqrt{\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; \&lsqb;</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; \&rsqb;</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; m</span>}}}$

(b) The current gas path health index of each component is obtained through the iterative optimization calculation of the particle swarm optimization algorithm, which is used to evaluate the actual component performance health status of the target gas turbine,

$\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; \&rsqb;</span>$

$\stackrel{<span\; class="notranslate">\; \&Right\; Arrow;</span>}{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; \&rsqb;</span>$

in Gas path measurement parameter vector calculated for the gas turbine thermal model; The measured gas path parameter vector; M is the number of gas path measurement parameters;

$\mathrm{<span\; class="notranslate">\; \&Delta;</span>}\stackrel{<span\; class="notranslate">\; \&Right\; Arrow;</span>}{\mathrm{<span\; class="notranslate">\; S</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; C</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; E.</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; E.</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; C</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; E.</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; \&rsqb;</span>$

in is the health index of gas circuit components;

Taking a root mean square error as the objective function, the following formula:

$\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; no</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; =</span>\sqrt{\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; \&lsqb;</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; \&rsqb;</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; m</span>}}}$

In the formula, Fitness is the optimization goal. With iterative optimization, when Fitness approaches 0, the calculated gas path measurement parameters and measured air path parameters match, and output the final global optimal solution at this time

The advantages of the present invention are:

(1) According to the aerothermodynamic characteristics of the gas turbine, the present invention uses the relative conversion parameters of the components to establish a gas turbine component-level thermal model, wherein the compressor and turbine characteristic line data are organized into a general relative conversion parameter form, compared with the prior art, it can It is easier and more convenient to set the health index of each gas circuit component when it is used for gas circuit diagnosis.

(2) The present invention redefines the gas path health index of the compressor and the turbine (representing the deviation of the characteristic line of the component) with relative conversion parameters, which can more accurately characterize the deterioration of the component performance compared with the prior art. The change of the gas path health index can eliminate the negative impact on the diagnosis result due to the change of environmental conditions (atmospheric pressure, temperature and relative humidity).

(3) From the perspective of global optimization (adopting particle swarm optimization algorithm), the present invention improves the characteristics of local optimization of traditional diagnostic algorithm (Newton-Raphson iterative algorithm), improves the accuracy of diagnostic results, and simplifies diagnosis The process can be effectively applied to the situation where there is measurement noise and performance diagnosis of complex gas turbine units.

Description of drawings

Fig. 1 is a flowchart of the present invention;

Fig. 2 is the schematic diagram of the diagnosis process of the present invention;

Figure 3 is an identification diagram of the working section of the gas circuit of a certain type of three-axis marine gas turbine;

Figure 4 is the diagnosis result of a certain type of three-shaft marine gas turbine diagnosis case;

Fig. 5 is the iterative calculation search process of the algorithm of the present invention in this diagnosis case.

detailed description

The present invention is described in more detail below in conjunction with accompanying drawing example:

In conjunction with Figures 1-5, the present invention is based on a thermodynamic model and a particle swarm optimization algorithm combined performance diagnosis method for gas turbine adaptive gas circuit components, including the following steps:

Step 1), based on the gas path measurement parameters when the target gas turbine is newly put into operation (or healthy), a non-linear thermal model of the gas turbine that can fully reflect the characteristics of each component is established, wherein both the compressor and the turbine are expressed in the form of similar converted parameters;

Step 2), redefine the gas path health index of the compressor and the turbine with similar conversion parameters, and eliminate the influence of the change in the operating performance of the gas turbine due to changes in environmental conditions (atmospheric pressure, temperature and relative humidity);

Step 3), collecting gas path measurement parameters of a certain period of time when the current target gas turbine is running stably, and performing noise reduction processing as gas path measurement parameters to be diagnosed offline;

Step 4), set the environmental input conditions (atmospheric pressure, temperature and relative humidity) and operating input conditions of the established gas turbine thermal model to be consistent with the operating conditions of the target gas turbine at the time of sampling, and eliminate due to environmental conditions (atmospheric pressure, temperature and relative humidity) Humidity) and changes in operating conditions that lead to changes in gas turbine performance.

Step 5), take the root mean square error between the gas path measurement data to be diagnosed offline and the gas path parameter data calculated by the thermal model as the objective function, and obtain the current components (compressor , turbine and combustor) gas path health index, which is used to evaluate the actual performance health status of the target gas turbine.

As a further optimization scheme of the gas turbine adaptive gas path component performance diagnosis method based on the combination of thermal model and particle swarm optimization algorithm in the present invention, the gas path measurement parameters when the target gas turbine is newly put into operation (or healthy) as described in step 1) The specific steps for establishing a nonlinear thermal model of a gas turbine that can fully reflect the characteristics of each component are as follows:

In step 1.1), according to the aerothermodynamic characteristics of the gas turbine, the component-level thermal model of the gas turbine is established by using the relative conversion parameters of the components, in which the compressor and turbine characteristic line data are organized into a common relative conversion parameter form.

Among them, the characteristic line of the compressor is sorted into a general similar conversion parameter form as follows:

G _C,cor,rel ＝f(n _cor,rel ,π _C,rel )

η _C,rel = f(n _cor,rel ,π _C,rel )

in is the relative converted speed, n is the actual speed, is the inlet stagnation temperature of the compressor, R _g is the gas constant of the working fluid flowing through the compressor, and the subscript 0 indicates the design point;

is the relative equivalent flow rate, G _C is the actual compressor inlet flow rate, is the compressor inlet stagnation pressure, is the relative pressure ratio, π _C is the actual compressor pressure ratio, η _C,rel = η _C /η _C0 is the relative isentropic efficiency, and η _C is the actual compressor isentropic efficiency.

The turbine characteristic line is sorted into a common similar converted parameter form as follows:

G _T,cor,rel ＝f(n _cor,rel ,π _T,rel )

η _T,rel = f(n _cor,rel ,π _T,rel )

In the formula: is the relative converted speed, n is the actual speed, is the stagnation temperature at the inlet of the turbine, R _g is the gas constant of the working fluid flowing through the turbine, is the relative converted flow, G _T is the actual turbine inlet flow, is the relative pressure ratio, η _T,rel = η _T /η _T0 is the relative isentropic efficiency, η _T is the actual turbine isentropic efficiency, and the subscript 0 indicates the design point.

Step 1.2), according to the gas path measurement parameters (after noise reduction treatment) collected when the target gas turbine is newly put into operation (or healthy), such as total temperature, total pressure, speed, etc., gradually correct the characteristic line data of each component (including design Working conditions and variable working conditions), so that the calculated value of the built gas turbine thermal model matches the measured parameters of the gas path, thereby eliminating the negative impact of the thermal model calculation error on the diagnosis results.

As a further optimization scheme of the gas turbine adaptive gas path component performance diagnosis method based on the combination of thermal model and particle swarm optimization algorithm in the present invention, the gas path health index of the compressor and the turbine is redefined with similar conversion parameters as described in step 2) The specific steps are as follows:

Step 2.1), the overall performance health of the gas turbine can be represented by the gas path health index of each main component, such as the flow characteristic index and efficiency characteristic index of the compressor and turbine, and the efficiency characteristic index of the combustor. Here, relative conversion parameters are used to redefine the gas path health index of the compressor and turbine (representing the deviation of the characteristic line of the component), in order to eliminate the diagnostic results due to changes in environmental conditions (atmospheric pressure, temperature and relative humidity) negative impact.

In step 2.2), the air path health index of the compressor is defined as follows:

SF _C,FC ＝G _{C,cor,rel,deg} /G _C,cor,rel

ΔSF _C,FC ＝(G _{C,cor,rel,deg} -G _C,cor,rel )/ _GC,cor,rel

SF _C,Eff =η _C,deg /η _C

ΔSF _C,Eff =(η _C,deg -η _C )/η _C

Among them, SF _{C, FC} is the flow characteristic index of the compressor; G _{C, cor, rel, deg} is the relative equivalent flow rate of the compressor (when the performance is declining); G _{C, cor, rel} is the relative equivalent flow rate of the compressor (when it is healthy); SF _{C, Eff} is the compressor efficiency characteristic index; η _{C, deg} is the isentropic efficiency of the compressor (when the performance is declining); η _C is the isentropic efficiency of the compressor (when it is healthy).

The combustion chamber gas path health index is defined as follows:

SF _B,Eff =η _B,deg /η _B

ΔSF _B,Eff =(η _B,deg -η _B )/η _B

Among them, SF _{B, Eff} is the combustion efficiency performance index of the combustion chamber; η _{B, deg} is the combustion efficiency of the combustion chamber (when the performance is declining); η _B is the combustion efficiency of the combustion chamber (when it is healthy).

The turbine gas path health index is defined as follows:

SF _T,FC ＝G _T,cor,deg /G _T,cor

ΔSF _T,FC ＝(G _T,cor,deg -G _T,cor )/G _T,cor

SF _T,Eff =η _T,deg /η _T

ΔSF _T,Eff =(η _T,deg -η _T )/η _T

Among them, SF _T,FC is the turbine flow performance index; G _T,cor,deg is the converted flow rate of the turbine (when the performance is declining); G _T,cor is the converted flow rate of the turbine (when it is healthy); SF _T,Eff is the turbine Efficiency performance index; η _T,deg is the isentropic efficiency of the turbine (when the performance is declining); η _T is the isentropic efficiency of the turbine (when it is healthy).

As a further optimization scheme of the gas turbine adaptive gas path component performance diagnosis method based on the combination of thermal model and particle swarm optimization algorithm in the present invention, the gas path calculated by the gas path measurement data to be diagnosed offline and the thermal model described in step 5) The root mean square error between the parameter data is the objective function, and the current gas path health index of each component (compressor, turbine, and combustor) is obtained through iterative optimization calculation of the particle swarm optimization algorithm, which is used to evaluate the actual performance of the target gas turbine The specific steps for health status are as follows:

Step 5.1) for gas path measurement data to be diagnosed offline Gas path parameter data calculated with thermal model The root mean square error between is the objective function Fitness:

$\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; no</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; =</span>\sqrt{\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; \&lsqb;</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; \&rsqb;</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; m</span>}}}$

In step 5.2), the current gas path health index of each component (compressor, turbine and combustion chamber) is obtained through iterative optimization calculation of the particle swarm optimization algorithm, which is used to evaluate the actual component performance health status of the target gas turbine. The schematic diagram of the gas turbine adaptive gas path component performance diagnosis method based on the combination of thermal model and particle swarm optimization algorithm is shown in Figure 2.

in,

$\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; \&rsqb;</span>$

$\stackrel{<span\; class="notranslate">\; \&Right\; Arrow;</span>}{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; \&rsqb;</span>$

in Gas path measurement parameter vector calculated for the gas turbine thermal model; The measured gas path parameter vector; M is the number of gas path measurement parameters;

$\mathrm{<span\; class="notranslate">\; \&Delta;</span>}\stackrel{<span\; class="notranslate">\; \&Right\; Arrow;</span>}{\mathrm{<span\; class="notranslate">\; S</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; C</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; E.</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; E.</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; C</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; E.</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; \&rsqb;</span>$

in - Gas path component health index (as a particle in the particle swarm optimization algorithm).

Here, a root mean square error is used as the objective function, such as:

$\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; no</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; =</span>\sqrt{\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; \&lsqb;</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; \&rsqb;</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; m</span>}}}$

In the formula, Fitness is the optimization goal. With iterative optimization, when Fitness gradually approaches 0, the calculated gas path measurement parameters and measured air path parameters match, and output the final global optimal solution at this time

The core technical content of the present invention is to solve the problem that the diagnostic accuracy of the traditional gas turbine gas circuit component performance diagnosis method is easily affected by the environmental conditions (atmospheric pressure, temperature and relative humidity) and the change of operating conditions, and from the perspective of global optimization, the traditional method is improved. The characteristic of local optimization of the diagnosis algorithm (Newton-Raphson iterative algorithm) improves the accuracy of the diagnosis result, simplifies the diagnosis process, and can be effectively applied to the performance diagnosis of the presence of measurement noise and complex gas turbine units. According to the aerothermodynamic characteristics of the gas turbine, the component-level thermal model of the gas turbine is established by using the relative conversion parameters of the components. Set the health index of each gas path component during gas path diagnosis; redefine the gas path health index of the compressor and turbine (representing the deviation of the characteristic line of the component) with relative conversion parameters, which can be more accurate than the existing technology Accurately characterize the change of gas path health index due to component performance degradation, and eliminate the negative impact on diagnosis results due to changes in environmental conditions (atmospheric pressure, temperature and relative humidity); from global optimization (using particle swarm optimization algorithm) It improves the local optimization characteristics of the traditional diagnostic algorithm (Newton-Raphson iterative algorithm), improves the accuracy of the diagnostic results, and simplifies the diagnostic process. It can be effectively applied to the performance of measurement noise and complex gas turbine units. Diagnose the situation.

Taking the health diagnosis of gas path components of a certain type of three-shaft marine gas turbine as an example, the identification diagram of the gas path working section of this type of three-shaft marine gas turbine is shown in Figure 3. This type of three-shaft gas turbine includes two compressors (a low-pressure compressor (LC) and a high-pressure compressor (HC)), a combustion chamber (B) and three turbines (a high-pressure turbine (HT), A low pressure turbine (LT) and a power turbine (PT)), where the generator is connected to the power turbine (PT) through a reduction gearbox. The output work of the low-pressure turbine (LT) drives the low-pressure compressor (LC) through the low-pressure shaft to compress the air coming out of the intake port, and the output work of the high-pressure turbine (HT) drives the high-pressure compressor (HC) through the high-pressure shaft to continue Compresses the air from the low pressure compressor (LC). The high-pressure air from the high-pressure compressor (HC) enters the combustion chamber (B) and undergoes a combustion chemical reaction with fuel to generate high-temperature and high-pressure gas, and the gas enters the high-pressure turbine (HT), low-pressure turbine (LT) and power turbine in sequence (PT) to drive the turbine output power. Ultimately, the power turbine (PT) drives a generator through a reduction gearbox to generate electrical power. At the same time, the cooling air extracted from the compressor flows into the hot-end airflow channel to cool the vanes, moving vanes and discs of the first few stages of each turbine. When the gas turbine is running stably, the electrical power of the generator and the rotational speed of the power turbine (PT) are usually kept constant as the main control parameters. The gas path measurement parameters of the unit are shown in Table 1, and the health index of each gas path component is shown in Table 2.

Table 1 Gas path measurement parameters of this type of gas turbine unit

Table 2 Health index of main gas circuit components

Based on the gas path measurement parameters of this type of gas turbine when it is newly put into operation (or healthy), a nonlinear thermal model of the gas turbine that can fully reflect the characteristics of each component is established, in which the compressor and turbine are expressed in the form of relative equivalent parameters.

Among them, the characteristic line of the compressor is sorted into a general similar converted parameter form as follows:

G _C,cor,rel ＝f(n _cor,rel ,π _C,rel )

η _C,rel = f(n _cor,rel ,π _C,rel )

in is the relative converted speed, n is the actual speed, is the inlet stagnation temperature of the compressor, R _g is the gas constant of the working fluid flowing through the compressor, and the subscript 0 indicates the design point;

is the relative equivalent flow rate, G _C is the actual compressor inlet flow rate, is the compressor inlet stagnation pressure, is the relative pressure ratio, π _C is the actual compressor pressure ratio, η _C,rel = η _C /η _C0 is the relative isentropic efficiency, and η _C is the actual compressor isentropic efficiency.

The turbine characteristic line is sorted into a common similar converted parameter form as follows:

G _T,cor,rel ＝f(n _cor,rel ,π _T,rel )

η _T,rel = f(n _cor,rel ,π _T,rel )

In the formula: is the relative converted speed, n is the actual speed, is the stagnation temperature at the inlet of the turbine, R _g is the gas constant of the working fluid flowing through the turbine, is the relative converted flow, G _T is the actual turbine inlet flow, is the relative pressure ratio, η _T,rel = η _T /η _T0 is the relative isentropic efficiency, η _T is the actual turbine isentropic efficiency, and the subscript 0 indicates the design point.

The thermal model of the target gas turbine is established on the Matlab simulation platform. The input conditions of the thermal model are environmental conditions (atmospheric temperature t ₀ , pressure P ₀ , relative humidity RH), generator output power Ne (as an operating condition), fuel composition, fuel lower calorific value, and gas path component health index (For newly put into operation units, ). The calculation output of the thermal model is the fuel flow G _f , the thermal parameters (such as total pressure and total temperature) and the rotational speed at the inlet and outlet sections of each component.

Redefine the gas path health index of compressors and turbines (representing the deviation of component characteristic lines) with relative conversion parameters to eliminate the influence of changes in gas turbine operating performance due to changes in environmental conditions (atmospheric pressure, temperature, and relative humidity).

The compressor gas path health index is defined as follows:

SF _C,FC ＝G _{C,cor,rel,deg} /G _C,cor,rel

ΔSF _C,FC ＝(G _{C,cor,rel,deg} -G _C,cor,rel )/ _GC,cor,rel

SF _C,Eff =η _C,deg /η _C

ΔSF _C,Eff =(η _C,deg -η _C )/η _C

Among them, SF _{C, FC} is the flow characteristic index of the compressor; G _{C, cor, rel, deg} is the relative equivalent flow rate of the compressor (when the performance is declining); G _{C, cor, rel} is the relative equivalent flow rate of the compressor (when it is healthy); SF _{C, Eff} is the compressor efficiency characteristic index; η _{C, deg} is the isentropic efficiency of the compressor (when the performance is declining); η _C is the isentropic efficiency of the compressor (when it is healthy).

The combustion chamber gas path health index is defined as follows:

SF _B,Eff =η _B,deg /η _B

ΔSF _B,Eff =(η _B,deg -η _B )/η _B

Among them, SF _{B, Eff} is the combustion efficiency performance index of the combustion chamber; η _{B, deg} is the combustion efficiency of the combustion chamber (when the performance is declining); η _B is the combustion efficiency of the combustion chamber (when it is healthy).

The turbine gas path health index is defined as follows:

SF _T,FC ＝G _T,cor,deg /G _T,cor

ΔSF _T,FC ＝(G _T,cor,deg -G _T,cor )/G _T,cor

SF _T,Eff =η _T,deg /η _T

ΔSF _T,Eff =(η _T,deg -η _T )/η _T

Among them, SF _T,FC is the turbine flow performance index; G _T,cor,deg is the converted flow rate of the turbine (when the performance is declining); G _T,cor is the converted flow rate of the turbine (when it is healthy); SF _T,Eff is the turbine Efficiency performance index; η _T,deg is the isentropic efficiency of the turbine (when the performance is declining); η _T is the isentropic efficiency of the turbine (when it is healthy).

The gas path measurement parameters of a certain period of time when the current target gas turbine is running stably are collected, and after noise reduction processing, they are used as the gas path measurement parameters to be diagnosed offline.

Set the environmental input conditions (atmospheric temperature t ₀ , pressure P ₀ , relative humidity RH) and operating input conditions (generator output power Ne) of the established gas turbine thermal model to be consistent with the operating conditions of the target gas turbine during sampling, eliminating the The effect of changes in gas turbine operating performance due to changes in operating conditions and operating conditions.

Gas path measurement data for offline diagnosis Gas path parameter data calculated with thermal model The root mean square error between is the objective function Fitness, and the current gas path health index of each component (compressor, turbine and combustor) is obtained through the iterative optimization calculation of the particle swarm optimization algorithm (for each component, it is used to evaluate the target gas turbine Actual component performance health status. The schematic diagram of the diagnosis process of the gas turbine adaptive gas path component performance diagnosis method based on the combination of thermal model and particle swarm optimization algorithm is shown in Figure 1.

in,

$\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; \&rsqb;</span>$

$\stackrel{<span\; class="notranslate">\; \&Right\; Arrow;</span>}{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; \&rsqb;</span>$

in Gas path measurement parameter vector calculated for the gas turbine thermal model; The measured gas path parameter vector; M is the number of gas path measurement parameters;

$\mathrm{<span\; class="notranslate">\; \&Delta;</span>}\stackrel{<span\; class="notranslate">\; \&Right\; Arrow;</span>}{\mathrm{<span\; class="notranslate">\; S</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; C</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; E.</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; E.</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; C</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; E.</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; \&rsqb;</span>$

in - Gas path component health index (as a particle in the particle swarm optimization algorithm).

Here, a root mean square error is used as the objective function, such as:

$\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; no</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; =</span>\sqrt{\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; \&lsqb;</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; \&rsqb;</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; m</span>}}}$

In the formula, Fitness is the optimization goal. With iterative optimization, when Fitness gradually approaches 0, the calculated gas path measurement parameters and measured air path parameters match, and output the final global optimal solution at this time

The settings of the relevant parameters of the particle swarm optimization algorithm are shown in Table 3, where the evolution algebra is 80 and the population size is 60, which is used to search for the optimal component health index (as shown in Table 3).

Table 3 Selection of relevant parameters of particle swarm optimization algorithm

parameter value population size 60 evolutionary algebra 80

After going through the diagnostic steps of the gas turbine adaptive gas path component performance diagnosis method based on the combination of the thermal model and the particle swarm optimization algorithm, the final diagnosis result can be obtained as shown in Fig. 4 .

Among them, PSO-GPA is the proposed gas turbine adaptive gas path component performance diagnosis method based on the combination of thermal model and particle swarm optimization algorithm, and GPA is the traditional diagnosis method. The iterative calculation search process of the gas turbine adaptive gas path component performance diagnosis method based on the combination of thermal model and particle swarm optimization algorithm in this diagnosis case is shown in Figure 5.

It can be seen from Figure 4 that since the core algorithm (Newton-Raphson iterative algorithm) of the traditional diagnostic method is essentially a local iterative optimization method, the gas turbine adaptive gas path components based on the combination of thermal model and particle swarm optimization algorithm The core algorithm of the performance diagnosis method is essentially a global iterative optimization method. Therefore, the gas turbine adaptive gas path component performance diagnosis method based on the combination of the thermal model and the particle swarm optimization algorithm can eliminate the fuzzy effect more effectively than the traditional diagnosis method. Accurately identify and isolate degraded components.

Claims (4)

1. A gas turbine adaptive gas path component performance diagnosis method based on the combination of thermal model and particle swarm optimization algorithm, which is characterized by: (1) Establish a nonlinear thermal model of the gas turbine based on the gas path measurement parameters when the target gas turbine is newly put into operation or healthy, in which the compressor and the turbine are expressed in the form of similar converted parameters; (2) Redefine the gas path health index of compressors and turbines with similar conversion parameters to eliminate the impact of changes in gas turbine operating performance due to changes in environmental conditions; (3) Collect gas path measurement parameters for a period of time when the current target gas turbine is running stably, and use it as gas path measurement parameters to be diagnosed offline after noise reduction processing; (4) Set the environmental input conditions and operating input conditions of the established gas turbine thermal model to be consistent with the operating conditions of the target gas turbine during sampling, and eliminate the influence of changes in the operating performance of the gas turbine due to changes in environmental conditions and operating conditions; (5) Taking the root mean square error between the gas path measurement parameters to be diagnosed offline and the gas path parameter data calculated by the thermal model as the objective function, the current gas path health of each component is obtained through iterative optimization calculation of the particle swarm optimization algorithm Index, used to evaluate the actual performance health of the target gas turbine. 2. the gas turbine adaptive gas path component performance diagnosis method based on thermal model and particle swarm optimization algorithm combined according to claim 1, is characterized in that: the concrete steps of setting up the nonlinear thermal model of gas turbine in the step (1) are as follows: (a) Establish a gas turbine component-level thermal model using the relative conversion parameters of the components, in which the compressor and turbine characteristic line data are organized into a common relative conversion parameter form: The characteristic line of the compressor is sorted into a common similar converted parameter form as follows: G _C,cor,rel ＝f(n _cor,rel ,π _C,rel ) η _C,rel = f(n _cor,rel ,π _C,rel ) in is the relative converted speed, n is the actual speed, is the inlet stagnation temperature of the compressor, R _g is the gas constant of the working fluid flowing through the compressor, and the subscript 0 indicates the design point; is the relative equivalent flow rate, G _C is the actual compressor inlet flow rate, is the compressor inlet stagnation pressure, is the relative pressure ratio, π _C is the actual compressor pressure ratio, η _{C, rel} = η _C / η _C is the relative isentropic efficiency, and η _C is the actual compressor isentropic efficiency; The turbine characteristic line is sorted into a common similar converted parameter form as follows: G _T,cor,rel ＝f(n _cor,rel ,π _T,rel ) η _T,rel = f(n _cor,rel ,π _T,rel ) In the formula: is the relative converted speed, n is the actual speed, is the stagnation temperature at the inlet of the turbine, R _g is the gas constant of the working fluid flowing through the turbine, is the relative converted flow, G _T is the actual turbine inlet flow, is the relative pressure ratio, ηT _,rel = _ηT / _ηT0 is the relative isentropic efficiency, _ηT is the actual turbine isentropic efficiency, and the subscript 0 indicates the design point; (b) According to the gas path measurement parameters collected when the target gas turbine is newly put into operation or healthy, gradually correct the characteristic line data of each component, so that the calculated value of the built gas turbine thermal model matches the measured parameters of the gas path, thereby eliminating the heat Negative impact of model calculation errors on diagnostic results. 3. the gas turbine adaptive gas path component performance diagnosis method based on thermodynamic model and particle swarm optimization algorithm combined according to claim 2, is characterized in that: step (2) redefines compressor and turbine with similar conversion parameters The specific steps of air path health index are as follows: (a) The overall performance and health status of the gas turbine is represented by the gas path health index of each component, and the gas path health index of the compressor and turbine is redefined with relative conversion parameters to eliminate the negative impact on the diagnosis results due to changes in environmental conditions. influences; (b) Compressor air path health index is defined as follows: SF _C,FC ＝G _{C,cor,rel,deg} /G _C,cor,rel ΔSF _C,FC ＝(G _{C,cor,rel,deg} -G _C,cor,rel )/ _GC,cor,rel SF _C,Eff =η _C,deg /η _C ΔSF _C,Eff =(η _C,deg -η _C )/η _C Among them, SF _{C, FC} is the flow characteristic index of the compressor; G _{C, cor, rel, deg} is the relative equivalent flow rate when the compressor performance declines; G _{C, cor, rel} is the relative equivalent flow rate when the compressor is healthy; SF _{C, Eff} is Compressor efficiency characteristic index; η _C,deg is the isentropic efficiency when the compressor performance declines; η _C is the isentropic efficiency when the compressor is healthy; The combustion chamber gas path health index is defined as follows: SF _B,Eff =η _B,deg /η _B ΔSF _B,Eff =(η _B,deg -η _B )/η _B Wherein SF _{B, Eff} is the combustion efficiency performance index of the combustion chamber; η _{B, deg} is the combustion efficiency when the performance of the combustion chamber declines; η _B is the combustion efficiency when the combustion chamber is healthy; The turbine gas path health index is defined as follows: SF _T,FC ＝G _T,cor,deg /G _T,cor ΔSF _T,FC ＝(G _T,cor,deg -G _T,cor )/G _T,cor SF _T,Eff =η _T,deg /η _T ΔSF _T,Eff =(η _T,deg -η _T )/η _T Among them, SF _T,FC is the turbine flow performance index; G _T,cor,deg is the equivalent flow rate when the turbine performance declines; G _T,cor is the equivalent flow rate when the turbine is healthy; SF _T,Eff is the turbine efficiency performance index; η _T,deg is the isentropic efficiency when the turbine performance declines; η _T is the isentropic efficiency when the turbine is healthy. 4. the gas turbine adaptive gas path component performance diagnosis method based on thermal model and particle swarm optimization algorithm combined according to claim 3, is characterized in that: step (5) is the gas path measurement parameter and thermal model to be diagnosed off-line The root mean square error between the calculated gas path parameter data is the objective function, and the current gas path health index of each component is obtained through the iterative optimization calculation of the particle swarm optimization algorithm, which is used to evaluate the actual performance and health status of the target gas turbine. as follows: (a) With gas path measurement data to be diagnosed offline Gas path parameter data calculated with thermal model The root mean square error between is the objective function Fitness: $\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; no</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; =</span>\sqrt{\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; \&lsqb;</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; \&rsqb;</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; m</span>}}}$ (b) The current gas path health index of each component is obtained through the iterative optimization calculation of the particle swarm optimization algorithm, which is used to evaluate the actual component performance health status of the target gas turbine, $\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; \&rsqb;</span>$ $\stackrel{<span\; class="notranslate">\; \&Right\; Arrow;</span>}{\mathrm{<span\; class="notranslate">\; z</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; ...</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; \&rsqb;</span>$ in Gas path measurement parameter vector calculated for the gas turbine thermal model; The measured gas path parameter vector; M is the number of gas path measurement parameters; $\mathrm{<span\; class="notranslate">\; \&Delta;</span>}\stackrel{<span\; class="notranslate">\; \&Right\; Arrow;</span>}{\mathrm{<span\; class="notranslate">\; S</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; C</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; E.</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; E.</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; C</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&Delta;SF</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; E.</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; \&rsqb;</span>$ in is the health index of gas circuit components; Taking a root mean square error as the objective function, the following formula: $\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; no</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; =</span>\sqrt{\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; \&lsqb;</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; p</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; u</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; \&rsqb;</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; m</span>}}}$ In the formula, Fitness is the optimization goal. With iterative optimization, when Fitness approaches 0, the calculated gas path measurement parameters and measured air path parameters match, and output the final global optimal solution at this time