CN108915900B - Fault Diagnosis Method of Liquid Rocket Engine Based on Mathematical Model Time-Invariant Information - Google Patents

Abstract

The invention provides a liquid rocket engine fault diagnosis method based on mathematical model time invariant information, which comprises the steps of firstly establishing mathematical models of main components of a liquid rocket engine and establishing time invariant coefficients representing the working state of the components; analyzing a change rule of a time invariant coefficient representing the state of each component influenced by the working state of the engine, and defining a threshold value of each component according to the change rule; for the liquid rocket engine to be detected, state data of all parts in the working state of the liquid rocket engine are collected, the time invariant coefficient of all the parts is calculated according to the collected state data and is compared with the threshold value determined in the front, and engine fault detection and diagnosis are carried out. The method solves the problem that the liquid rocket engine is difficult to detect and diagnose under the conditions of lack of prior knowledge, insufficient fault samples, incomplete fault modes and the like, and can effectively realize the fault detection and diagnosis of the liquid rocket engine under the difficult conditions.

Description

Fault Diagnosis Method of Liquid Rocket Engine Based on Mathematical Model Time-Invariant Information

technical field

The invention relates to the field of liquid rocket engine fault detection, in particular to a liquid rocket engine fault diagnosis method based on mathematical model time-invariant information.

Background technique

The liquid rocket engine is the main power device and key component of the launch vehicle propulsion system, but at the same time, the liquid rocket engine is a very complex fluid-thermal power system, which not only works in harsh environments such as high temperature, high pressure, strong vibration and strong corrosion, but also Moreover, the energy release in the working stage is very concentrated, so it is a frequent failure site. So far, the world has been studying liquid rocket engines for nearly a hundred years, and various fault detection and isolation technologies are becoming more and more mature. However, in actual launch missions, losses and disasters caused by engine failures are still unavoidable. Once a rocket engine fails, it will affect the performance of the engine at least, and lead to the failure of space missions and even endanger the lives of astronauts, causing inestimable losses. On July 26, 2006, a carrier rocket transformed from a Russian-made RS-20 heavy intercontinental ballistic missile carrying 18 satellites was launched at the Bekonor Space Center in Kazakhstan, and crashed due to engine failure shortly after lift-off; in 2010 On December 25, the GSLV-F06 carrier rocket carrying the Indian-made GSAT-5P satellite began to smoke and deviated from orbit less than 1 minute after launch due to a serious technical failure of the first-stage engine, about 19 minutes Afterwards, the rocket exploded violently in the air, destroying both the star and the arrow; on August 24, 2011, the "Union-U" rocket carrying the "Progress M-12M" cargo spacecraft exploded shortly after lift-off. , It was found that the power equipment of the third stage of the rocket had failed; on December 23, 2011, Russia launched the "Soyuz-2.1B" rocket carrying the "Meridian" communication satellite, but failed due to the failure of the third stage rocket engine. Entered the scheduled orbit; on July 2, 2013, the Russian "Proton M/DM3" carrier rocket launched at the Baikonur launch site 17 seconds after ignition, the first-stage booster suddenly entered the failure mode and shut down, causing the rocket to upload Some 600 tons of toxic fuel (unsymmetrical dimethylhydrazine) leaked, causing large-scale local environmental pollution; on May 22, 2014, the AJ-26 liquid oxygen kerosene engine that was tested at the Stennis Space Center in the United States was ignited for 30 seconds A failure occurred at the time, causing extensive damage to the engine, and the test was forced to stop; on October 28, 2014, when the "Antares 2" carrier rocket was launched at the Wallops Flight Center in the United States, due to a failure of the rocket engine, it took off within 6 hours of ignition. Seconds later, it crashed on the launch site, and the "Cygnus" cargo spacecraft carried by the rocket suffered heavy losses; The failure of the third-stage engine caused the rocket to crash; on June 28, 2015, the "Dragon" spacecraft, which planned to deliver a large amount of supplies to the International Space Station, was launched from Canavera, Florida by the "Falcon 9" carrier rocket developed by SpaceX of the United States. Cape Earl Air Force Base launched, but only a few minutes after launch, the rocket engine failed and exploded in the air. Another launch accident of "Falcon 9" was on October 7, 2012, also due to One of the rocket's engines failed, causing an OG2 prototype communications satellite it carried to fail to reach its intended orbit. According to statistics, although space powers (such as Russia, the United States, and China) have launched many rockets, they have also failed many times. The highest success rate is only 96.1%, and India and Israel are relatively low, with a success rate of only 100%. Sixty or seventy. Therefore, it is of great theoretical significance and engineering practical value to carry out research on liquid rocket engine fault detection and diagnosis methods.

The current research on liquid rocket engine fault detection and diagnosis methods mainly includes three categories: methods based on test signal data statistics, methods based on mathematical models and methods based on artificial intelligence. The method based on data statistics relies on enough data samples, and obtains the relevant laws of the engine's working state through the statistical analysis of the data samples, so as to determine the judgment threshold of the relevant measurement parameters, and then judge the engine or components according to certain threshold inspection rules Whether there is a fault; the method based on artificial intelligence has unique advantages in dealing with fault detection and diagnosis of complex systems, but the premise also requires a large number of data samples for training. For liquid rocket engines with few fault samples, or new liquids with few normal samples Rocket engines, neither of these two types of methods can adapt. There are mainly two types of methods based on mathematical models: quantitative mathematical models and qualitative mathematical models. The method based on quantitative mathematical model is effective in dealing with linear systems and is widely used, but for complex liquid rocket engine systems, it is usually difficult to establish an accurate mathematical model, which limits the application of this method in engine fault detection and diagnosis; based on qualitative mathematical model The method will generate a lot of false information while getting the real solution, so the accuracy of fault diagnosis is not very high.

Therefore, it remains to develop a fault detection and diagnosis method suitable for liquid rocket motor systems.

Contents of the invention

The present invention is aimed at fault detection and detection due to the lack of prior knowledge, insufficient fault samples, and incomplete fault modes, especially for new liquid rocket engines that are still in the trial production stage, because not only the fault samples are lacking, but the normal samples are also very limited. For the problem of difficult diagnosis, a liquid rocket engine fault diagnosis method based on the time-invariant information of the mathematical model is proposed. This method is an envelope method using the time-invariant information in the mathematical model of the engine system as a statistical index.

For realizing above-mentioned technical purpose, technical scheme of the present invention is:

A liquid rocket engine fault diagnosis method based on mathematical model time-invariant information, comprising the following steps:

S1: For the liquid rocket engine, establish a mathematical model of its main components.

S2: Based on the mathematical model of each component of the liquid rocket engine, construct the time-invariant coefficients in each mathematical model that can represent the normal or fault state of the component.

S3: Analyzing the variation rules of the time-invariant coefficients representing the state of each component affected by the working state of the engine, and defining the thresholds of each time-invariant coefficient according to the variation rules.

S4: For the liquid rocket engine to be tested, collect the state data of each component in its working state, calculate the time-invariant coefficient of each component according to the collected state data and compare it with the threshold determined in S3, and perform engine failure Detection and diagnosis. If all the time-invariant coefficients are within the threshold range, it is judged that the engine is normal; if there is a time-invariant coefficient representing the state of a certain component that exceeds the threshold for w consecutive times, it is considered that the component of the engine is faulty, so as to realize engine fault detection and diagnosis .

For liquid rocket engines, its main components include pumps, gas turbines, thermal components and liquid pipelines; pumps include oxidizer pumps and fuel pumps; gas turbines include fuel turbines and oxidizer turbines; thermal components include gas generators, combustion chambers and gas Conduits; liquid pipelines including propellant delivery pipelines, etc.

Therefore, in S1, the corresponding mathematical models for the pump, gas turbine, thermal components and liquid pipelines of the liquid rocket engine were constructed respectively.

Construct the corresponding mathematical model for the pump of the liquid rocket engine as the pump model, as follows:

Among them, ΔP is the head of the pump, P _pe and P _pi represent the outlet and inlet pressure of the pump respectively, n _p is the speed of the pump, q _p is the flow rate of the pump, and μ _p1 , μ _p2 , μ _p3 respectively represent the experience of the pump head coefficient.

The corresponding mathematical model for the gas turbine of the liquid rocket engine is the gas turbine model, as follows:

Gas turbine power equation:

Where gas turbine efficiency η:

Among them, n is the gas turbine speed, b ₁ , b ₂ , b ₃ are empirical coefficients.

Gas turbine flow q _t :

Among them, k, R, and T are the adiabatic index, gas constant, and temperature of the gas, respectively; the subscript i in (RT) _i represents the inlet, which is the inlet of the gas turbine here; q _t represents the flow of gas flowing through the gas turbine, P _t0 is the gas pressure at the inlet of the gas turbine, P _te is the gas pressure at the outlet of the gas turbine, P _ti is the gas pressure between the stator and the rotor of the gas turbine, θ is the reaction force, μ is the flow coefficient of the gas turbine nozzle; A _t is the flow coefficient of the gas turbine The area of the nozzle.

Gas turbine theoretical injection velocity V _e :

Gas turbine power balance equation:

Among them, N is the gas turbine power; ΣN _p represents the sum of the pump power driven by the gas turbine, and J represents the moment of inertia of the gas turbine pump rotor.

The corresponding mathematical model for the thermal components of the liquid rocket engine is constructed as a thermal component model, as follows:

The mass conservation equation in the thermodynamic component:

Calculation formula of gas density change law in thermal components:

The rate of change of the gas mixture ratio in the thermal assembly:

Among them, q _o represents the oxidant flow rate, and q _f represents the fuel flow rate.

The calorific value of gas is calculated according to the difference of the mixing ratio:

RT=RT(r)

Among them, T(r) means that T is a function of the gas mixture ratio r, and changes with the gas mixture ratio r.

According to the ideal gas equation of state

PV = m _g RT

Carrying out derivation processing, we can get

The outlet flow equation can then be analyzed

Among them, m _g , _ρ , V, _P and _r are the gas mass, density, volume, pressure and mixing ratio in the thermal component respectively; Mass flow rate and liquid fuel mass flow rate; q _eg is the outlet flow rate of the thermal component; ζ is the flow coefficient of the throat of the thermal component; A is the throat area of the thermal component.

Constructing the corresponding mathematical model for the liquid pipeline of the liquid rocket engine is a liquid pipeline model, as follows:

The flow equation of the liquid propellant in the liquid pipeline is as follows:

The continuity equation of the propellant components in the liquid pipeline is as follows:

Among them, α, ξ, and L are the flow resistance coefficient, flow capacity coefficient and inertial flow resistance coefficient of the liquid pipeline respectively; q _li , P _li , q _le and P _el represent the mass of the inlet and outlet of the liquid pipeline flow and pressure; V _l is the volume of the liquid pipeline; a represents the sound velocity in the liquid in the liquid pipeline.

In S2 of the present invention, the time-invariant coefficient of the characterizing pump working state of construction is:

The time-invariant coefficient constructed to characterize the working state of the gas turbine is:

The time-invariant coefficient constructed to characterize the working state of the thermal component is:

The time-invariant coefficient constructed to characterize the working state of the liquid pipeline is:

Among them, P _le and P _li are the outlet and inlet pressures of the liquid pipeline, respectively, and q _l is the flow rate of the liquid pipeline.

In S3 of the present invention, the sample library is first established, and the state data of each component under the working state of multiple liquid rocket engines of the same model is collected for a period of time, and the time-invariant coefficient of each component of the liquid rocket engine is calculated according to the collected state data. , the change law of the time-invariant coefficient of the state of each component of the liquid rocket engine affected by the working state of the engine is obtained statistically, and the threshold value of each time-invariant coefficient is defined according to the change law.

The more sample data in the sample library, the more accurate the change law of the time-invariant coefficient of the state of each component of this type of liquid rocket engine affected by the working state of the engine will be obtained from the final statistics.

When the engine is working normally, the time-invariant coefficient of each component state corresponds to a value or an interval; when the engine is in a different fault state or the fault degree becomes larger or smaller, the time-invariant coefficient of each component state will occur accordingly Change and deviate from the normal value or normal interval, these changing rules are the changing rules that the time-invariant coefficients of the state of each part of each liquid rocket engine obtained by the statistical method are affected by the working state of the engine. The method of defining the threshold according to the above-mentioned change law obtained by statistics includes numerical features such as expectation and variance commonly used in mathematical statistics, or various estimation methods such as point estimation and interval estimation.

In S4 of the present invention, w is a pre-set integer greater than 1, and its specific value is generally set according to actual conditions and experience, generally three times.

Compared with the prior art, the present invention can produce the following technical effects:

The present invention provides a convenient and reliable method for the problem of difficulty in detecting and diagnosing liquid rocket engine faults under the conditions of lack of prior knowledge, insufficient fault samples, and incomplete fault modes, etc., which can effectively realize the faults of liquid rocket motors under the aforementioned difficult conditions Detection and diagnosis.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. Those skilled in the art can also obtain other drawings based on these drawings without creative work.

Fig. 1 is the exterior view of liquid rocket engine of the present invention;

Figure 2 is a hierarchical exploded view of the liquid rocket engine system of the present invention;

Fig. 3 is the implementation mode and step diagram of fault detection and diagnosis of the present invention;

Fig. 4 is the fault detection and diagnosis result of the embodiment of the present invention;

Detailed ways

The embodiments of the present invention will be described in detail below with reference to the accompanying drawings, but the present invention can be implemented in many different ways defined and covered by the claims.

The basic thinking of the present invention is: carry out modular decomposition to liquid rocket engine, and establish the mathematical model of each component respectively, construct the time-invariant coefficient that characterizes component normal or fault state in each modularized mathematical model, and analyze these time-invariant Coefficients vary with the state of the system; when engine components fail, the time-invariant coefficients of these components corresponding to the mathematical model will change, thereby affecting the change of its output state; in this way, the component can be represented by each modular mathematical model The time-invariant coefficient of the state is detected in real time, and the fault detection and diagnosis of the liquid rocket engine can be realized according to whether the statistical threshold of its change law is exceeded.

In order to illustrate the technical solution of the present invention, the following will be described through specific embodiments in conjunction with the accompanying drawings.

The technical solution of this embodiment is: a fault detection and diagnosis method based on the time-invariant information of the mathematical model of the liquid rocket engine system, for the engine shown in Figure 1, according to Figure 2, the liquid rocket engine hierarchy is divided into engine levels , subsystem level and component level, etc., and establish mathematical models of each component. Then carry out the following steps: analyze the mathematical model of each component one by one, construct a time-invariant coefficient that can characterize the normal or fault state of the component, and analyze its variation rule affected by the engine state during the engine working process. This layered structure is beneficial to distinguish component failures. When the engine fails, the fault can be detected according to the fault state displayed by the engine, and then the time-invariant coefficient changes that characterize the normal or fault state of each component can be analyzed in turn to find the faulty component. Find out the cause of the fault and complete fault detection and diagnosis.

As shown in Figure 1 and Figure 2, this method is suitable for liquid rocket engine fault detection and diagnosis. The liquid rocket engine is composed of thrust chamber, gas generator, turbo pump, propellant supply system, valves and regulating components. The engine system is divided into engine level, subsystem level and component level according to the level. The subsystem level includes thrust chamber system, turbo pump system, pipeline system and gas generator system. The subsystem can be further subdivided into component level, the thrust chamber system can be subdivided into nozzle, combustion chamber, injector head, etc.; the turbopump system can be further divided into fuel turbine and fuel pump, oxidizer turbine and oxidant pump; The pipeline subsystem can be further divided into liquid pipelines, gas pipelines, throttle valves, flow regulators, pipelines with valves, etc. Since the similar parts of the engine have the same mathematical form, it is given in a modular form in the modeling, and the analysis can be carried out by calling the module for different engines.

With reference to Fig. 3, a kind of liquid rocket engine fault diagnosis method based on mathematical model time-invariant information provided by the present invention in detail below, comprises the following steps:

S1: For the liquid rocket engine, establish a mathematical model of its main components.

S2: Based on the mathematical model of each component of the liquid rocket engine, construct the time-invariant coefficient in each mathematical model that can represent the normal or fault state of the component;

Different liquid rocket engines share many similar major components, including pumps, turbines, thermal components, liquid lines, and more. The pumps include oxidant pumps and fuel pumps; turbines include fuel turbines and oxidizer turbines; thermal components include gas generators, combustion chambers and gas conduits; liquid pipelines include propellant delivery pipelines, etc.

Similar components have many differences in details, but generally have the same mathematical form, so the modeling is given in a modular form. For different liquid rocket engines, they can be analyzed by calling their component modules. Next, the mathematical model of the main components is established, and the time-invariant coefficients representing the components are constructed.

(1) The pump model is suitable for various pumps in liquid rocket engines, such as oxidant pumps and fuel pumps.

The parameters that characterize the main performance of the pump are flow, head, speed, power and efficiency.

Pump head: The energy increase value of each unit mass of propellant passing through the pump is called the pump head.

Among them, ΔP is the head of the pump, P _pe and P _pi represent the outlet and inlet pressure of the pump respectively, n _p is the speed of the pump, q _p is the flow rate of the pump, and μ _p1 , μ _p2 , μ _p3 respectively represent the experience of the pump head The coefficient is a known parameter, and the empirical parameter of the batch of products is determined by the manufacturer of the pump after production through tests and other methods.

The so-called time-invariant coefficient here means that under normal working conditions, the coefficient does not change with time. The time-invariant coefficient always corresponds to a constant or an interval. When the system fails or deviates from the normal working condition, the coefficient will deviate from the constant or interval.

After analysis, the time-invariant coefficient to characterize the working state of the pump is constructed as follows:

The oxidizer pump is the same as the fuel pump.

(2) Gas turbine model

Next build the gas turbine model.

Gas turbine power equation:

The power of the turbine is determined by the power required by the pump.

Gas turbine efficiency equation:

Among them, b ₁ , b ₂ , and b ₃ are empirical coefficients, which are known parameters, and the manufacturer of the gas turbine shall determine the empirical parameters of this batch of products through tests and other methods after production.

Gas turbine flow equation:

Among them, k, R, and T are the adiabatic index, gas constant, and temperature of the gas, respectively; q _t represents the gas flow through the gas turbine, P _t0 is the gas pressure at the gas turbine inlet, P _te is the gas pressure at the gas turbine outlet, and P _ti is the gas pressure between the gas turbine stator and rotor, θ is the reaction force, μ is the flow coefficient of the gas turbine nozzle; _At is the area of the gas turbine nozzle.

Gas Turbine Theoretical Injection Velocity:

Gas turbine power balance equation:

Among them, N is the gas turbine power; ΣN _p represents the sum of the pump power driven by the gas turbine, J represents the moment of inertia of the gas turbine pump rotor, and n is the speed of the gas turbine.

After analysis, the time-invariant coefficients constructed to characterize the working state of the gas turbine are:

Oxidant turbines and fuel turbines are both gas turbines. The gas turbine model constructed above and the time-invariant coefficients representing the working state of the gas turbine are applicable to the oxidant turbine and the fuel turbine

(3) Thermal component model

Thermal components include gas generators, combustion chambers and gas ducts.

The mass conservation equation in the thermodynamic component:

Calculation formula of gas density change law in thermal components:

The rate of change of the gas mixture ratio in the thermal assembly:

Among them, q _o represents the oxidant flow rate, and q _f represents the fuel flow rate.

The calorific value of gas is calculated according to the difference of the mixing ratio:

RT=RT(r)

Among them, T(r) means that T is a function of the gas mixture ratio r, and changes with the gas mixture ratio r. This function is given as a known function, and the changing function is generally obtained by fitting the manufacturer of the thermal component according to experience and experimental data.

According to the ideal gas equation of state

PV = m _g RT

Carrying out derivation processing, we can get

The outlet flow equation can then be analyzed

Among them, m _g , _ρ , V, _P and _r are the gas mass, density, volume, pressure and mixing ratio in the thermal component respectively; Mass flow rate and mass flow rate of liquid fuel; q _eg is the outlet flow rate of the thermal component; ζ is the flow coefficient of the throat of the thermal component (that is, the position where the cross section of the thermal component is the smallest); A is the throat area of the thermal component.

In the present invention, the time-invariant coefficient constructed to characterize the working state of the thermal component is:

The thermal component model constructed above and the time-invariant coefficients representing the working state of the thermal component are applicable to the gas generator, combustion chamber and gas conduit.

(4) Liquid pipeline model

The flow equation of the liquid propellant in the liquid pipeline is as follows:

The continuity equation of the propellant components in the liquid pipeline is as follows:

Among them, α, ξ, and L are the flow resistance coefficient, flow capacity coefficient and inertia flow resistance coefficient of the liquid respectively; q _li , P _li , q _le and P _el respectively represent the mass of the inlet and outlet of the liquid pipeline flow and pressure. V _l is the volume of the liquid pipeline; a represents the sound velocity in the liquid in the liquid pipeline.

To construct the time-invariant coefficient representing the working state of the liquid pipeline is:

Among them, P _le and P _li are the outlet and inlet pressures of the pipeline respectively, and q _l is the flow rate of the pipeline.

The liquid pipeline model constructed above and the time-invariant coefficients characterizing the working state of the liquid pipeline are applicable to all propellant delivery pipelines, including pipelines with valves.

S3: Analyzing the change law of the time-invariant coefficients that characterize the state of each component affected by the working state of the engine, and defining the threshold of each time-invariant coefficient according to the change law;

For S3, a sample library should be established first, and the state data of various components of this type of liquid rocket engine under the working state should be collected for a period of time. According to the collected state data, the time-invariant coefficient of each component of this type of liquid rocket engine should be calculated. The change law of the time-invariant coefficient of the state of each component of the liquid rocket engine affected by the working state of the engine is obtained statistically, and the threshold value of each time-invariant coefficient is defined according to the change law.

The more sample data in the sample library, the more accurate the change law of the time-invariant coefficient of the state of each component of this type of liquid rocket engine affected by the working state of the engine will be obtained from the final statistics.

When the engine is working normally, the time-invariant coefficient of each component state corresponds to a value or an interval; when the engine is in a different fault state or the fault degree becomes larger or smaller, the time-invariant coefficient of each component state will occur accordingly Change and deviate from the normal value or normal interval, these changing rules are the changing rules that the time-invariant coefficients of the state of each part of each liquid rocket engine obtained by the statistical method are affected by the working state of the engine. The method of defining the threshold according to the above-mentioned change law obtained by statistics includes numerical features such as expectation and variance commonly used in mathematical statistics, or various estimation methods such as point estimation and interval estimation.

S4: For the liquid rocket engine to be tested, collect the state data of each component in its working state, calculate the time-invariant coefficient of each component according to the collected state data and compare it with the threshold determined in S3, and perform engine failure Detection and diagnosis. If all the time-invariant coefficients are within the threshold range, it is judged that the engine is normal; if there is a time-invariant coefficient representing the state of a certain component that exceeds the threshold for w consecutive times, it is considered that the component of the engine is faulty, so as to realize engine fault detection and diagnosis . Among them, w is a pre-set integer greater than 1, which is set according to the actual situation and experience, generally 3 times.

The following describes the fault detection and diagnosis of the fuel cold exhaust cavitation pipe of a certain type of liquid rocket engine.

According to the components of the engine system, the time-invariant coefficients for characterizing each component are constructed according to the method of the present invention, and their variation rules are analyzed. Define its threshold, and then perform fault detection and diagnosis.

Figure 4 shows the fault detection and diagnosis results. It can be seen from Figure 4 that the time-invariant coefficient representing the state of the fuel cold exhaust cavitation tube exceeds the threshold at about 430s, while the other time-invariant coefficients have not changed significantly, so the fuel cold exhaust can be detected and diagnosed Corrosion tube failed.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims (8)

1. it is a kind of based on mathematical model when fixed information Liquid Propellant Rocket Engine Fault Diagnosis method, it is characterised in that: including Following steps:

S1: it is directed to liquid-propellant rocket engine, establishes the mathematical model of its each main building block；

S2: the mathematical model based on each building block of liquid-propellant rocket engine, the component can be characterized by constructing in each mathematical model Normal or malfunction when constant coefficient；

S3: it is analyzed and characterized the changing rule that the when constant coefficient of each unit status is influenced by engine behavior, and according to institute State the threshold value of constant coefficient when changing rule defines each；Constant coefficient refers to when liquid-propellant rocket engine normal operation when described Under, the when constant coefficient for characterizing each unit status does not change over time, and the when constant coefficient for characterizing each unit status corresponds to always One constant or a section characterize each unit status when liquid-propellant rocket engine breaks down or deviates nominal situation When constant coefficient will deviate from the constant or section；

S4: for liquid-propellant rocket engine to be detected, acquiring the status data of each component under its working condition, according to collecting Status data constant coefficient and itself and the threshold value that determines in S3 are compared when calculating each component, carry out engine failure Detection and diagnosis；If institute sometimes constant coefficient all in threshold range, judging that engine is normal；If there is characterize some When constant coefficient continuous w times of unit status exceeds threshold value, then it is assumed that the engine component malfunction, to realize engine Fault detection and diagnosis.

2. it is according to claim 1 based on mathematical model when fixed information Liquid Propellant Rocket Engine Fault Diagnosis method, It is characterized by:

Its main building block of liquid-propellant rocket engine includes pump, gas turbine, heating power component and liquid line；Wherein pump includes Oxidant pump, petrolift；Gas turbine includes fueled turbine, oxidant turbine；Heating power component includes gas generator, combustion chamber And gas conduct pipe；Liquid line includes propellant energy properties pipeline；

In S1, its corresponding number is constructed respectively for the pump of liquid-propellant rocket engine, gas turbine, heating power component and liquid line Learn model.

3. it is according to claim 2 based on mathematical model when fixed information Liquid Propellant Rocket Engine Fault Diagnosis method, It is characterized by:

In S1, for liquid-propellant rocket engine pump construct its corresponding mathematical model for pump model, it is as follows:

Wherein, Δ P is the lift of pump, P_pe、P_piRespectively indicate the outlet and inlet pressure of pump, n_pFor the revolving speed of pump, q_pFor pump Flow, μ_p1、μ_p2、μ_p3Respectively indicate the empirical coefficient of pump lift；

In S1, for liquid-propellant rocket engine gas turbine construct its corresponding mathematical model be gas turbine model, it is as follows:

Gas turbine power equation:

Wherein turbine efficiency η:

Wherein, n is gas turbine revolving speed, b₁, b₂, b₃For empirical coefficient；

Gas turbine flow q_t:

Wherein, k, R, T are respectively the adiabatic exponent, gas constant, temperature of combustion gas；q_tThe gas flow for flowing through gas turbine is represented, P_t0For gas turbine entrance gaseous-pressure, P_teFor gas turbine outlet gas pressure, P_tiBetween gas turbine stator and rotor Gaseous-pressure, θ are reaction degree, and μ is the discharge coefficient of gas turbine nozzle；A_tFor the area of gas turbine nozzle；

Gas turbine theory jet velocity V_e:

Gas turbine power equilibrium equation:

Wherein N is gas turbine power；∑N_pIndicate that the sum of the pump power driven by gas turbine, J indicate gas turbine pump rotor Rotary inertia；

It is heating power component model for its corresponding mathematical model of the heating power component construction of liquid-propellant rocket engine in S1, as follows:

Mass-conservation equation in heating power component:

Combustion gas density change calculating formula in heating power component:

The change rate of combustion gas mixing ratio in heating power component:

Wherein, q_oRepresent oxidizer flow rate, q_fRepresent fuel flow rate；

Fuel gases calorific value carries out difference calculating according to mixing ratio:

RT=RT (r)

Wherein, T (r) indicates that T is the function of combustion gas mixing ratio r, as combustion gas mixing ratio r changes；

According to The Ideal-Gas Equation

PV=m_gRT

Derivation processing is carried out, can be obtained

And then it can analyze rate of discharge equation

Wherein, m_g, ρ, V, P and r be respectively combustion gas quality, density, volume, pressure and mixing ratio in heating power component；q_ig、q_loWith q_lfRespectively flow into combustion gas quality flow, liquid oxidizer mass flow and the liquid fuel mass flow of heating power component；q_egFor The rate of discharge of heating power component；ζ is the discharge coefficient of the throat of heating power component；A is the throat opening area of heating power component；

In S1, for liquid-propellant rocket engine liquid line construct its corresponding mathematical model be liquid line model, it is as follows:

The flow equation of liquid propellant such as formula in liquid line:

The continuity equation of propellant constituent element such as formula in liquid line:

Wherein α, ξ and L are respectively the inertia flow resistance coefficient of the flow resistance coefficient of liquid line, fluid capacitance coefficient and liquid；q_li、P_li、 q_leAnd P_elRespectively indicate the entrance of liquid line, the mass flow of outlet and pressure；V_lFor liquid line volume；a Indicate the velocity of sound in liquid line in liquid.

4. it is according to claim 3 based on mathematical model when fixed information Liquid Propellant Rocket Engine Fault Diagnosis method, It is characterized by: in S2, the when constant coefficient of the characterization pump work state of building are as follows:

In S2, the when constant coefficient of the characterization gas turbine operation state of building are as follows:

In S2, the when constant coefficient of the characterization heating power component operation state of building are as follows:

In S2, the when constant coefficient of the characterization liquid line working condition of building are as follows:

Wherein, P_leAnd P_liThe respectively outlet and inlet pressure of liquid line, q_lFor piping flow.

5. it is according to claim 4 based on mathematical model when fixed information Liquid Propellant Rocket Engine Fault Diagnosis method, It is characterized by: first establishing sample database in S3, the work shape of more liquid-propellant rocket engines of same model in a period of time is acquired The status data of each component under state, according to collected status data, when calculating each component of the model liquid-propellant rocket engine not Variable coefficient, statistics obtain the variation that the when constant coefficient of each unit status of liquid-propellant rocket engine is influenced by engine behavior Rule, and when defining each according to the changing rule constant coefficient threshold value.

6. it is according to claim 5 based on mathematical model when fixed information Liquid Propellant Rocket Engine Fault Diagnosis method, It is characterized by: in S3, it is more for the sample data volume in sample database, it finally counts the obtained model liquid rocket and starts The changing rule that the when constant coefficient of each unit status of machine is influenced by engine behavior will be more accurate；

In engine work, when constant coefficient corresponding one of each unit status is worth or a section；Work as hair When motivation is in different malfunction or fault degree and becomes larger or become smaller, each unit status when constant coefficient can be corresponding It changes and deviates normal value or normal interval, these count obtained each liquid-propellant rocket engine its each unit status When the changing rule that is influenced by engine behavior of constant coefficient.

7. it is according to claim 1 based on mathematical model when fixed information Liquid Propellant Rocket Engine Fault Diagnosis method, It is characterized by: in S4, w be it is pre-set be greater than 1 integer.

8. it is according to claim 7 based on mathematical model when fixed information Liquid Propellant Rocket Engine Fault Diagnosis method, It is characterized by: w is set as 3 in S4.