Sensor optimization layout method and device for online condition monitoring of turbine cooler

Abstract

The present application provides a method and device for optimizing the layout of sensors for online status monitoring of a turbine cooler, which belongs to the field of aviation electromechanical. The method provided in the present application includes determining the sensor type based on the fault monitoring signal type and fault coverage index of the turbine cooler; establishing a turbine cooler simulation model; obtaining a fault signal, and inputting the fault signal into the simulation model, wherein the fault signal includes at least operating parameters and vibration parameters; identifying fault-sensitive monitoring points based on the correlation analysis between the output result of the simulation model and the fault signal, and determining the sensor layout position; arranging the online status monitoring sensor of the turbine cooler based on the sensor layout position and the sensor type. The method and device for optimizing the layout of sensors for online status monitoring of a turbine cooler provided in the present application can monitor the operating status of a turbine cooler online, obtain fault information in real time, and achieve high-precision fault location.

Description

Technical Field

[0001] The present application relates to the field of aviation electromechanical technology, and in particular to a sensor optimization layout method and device for online status monitoring of a turbine cooler. Background Art

[0002] The turbine cooler is a core component of the aircraft environmental control system. It uses the compressed hot air from the engine compressor. After being cooled in the front stage, it enters the fuel-air heat exchanger and then flows into the nozzle ring along the turbine cooler volute. The high-speed air flows into the turbine disc blades from the nozzle ring, impacts the impeller to rotate and do work, the air expands, the pressure decreases, the temperature drops, and the kinetic energy and enthalpy drop of the air are converted into mechanical work, which is consumed by the fan. After the air works on the blades of the turbine disc, it leaves the turbine cooler at low pressure and low temperature and is supplied to the aircraft cabin and instrument compartment, while part of the air is used to cool the bearings of the turbine cooler.

[0003] As a high-speed rotating machine, turbine coolers often experience problems such as jamming, breakage, and abnormal wear during testing and use. Once a failure occurs, it will lead to serious consequences and even affect flight safety. How to detect product operation abnormalities at an early stage and diagnose the fault is of great significance to improving product quality.

[0004] At present, aircraft turbine coolers are still mainly pure mechanical products, with relatively scarce monitoring methods, and are still mainly maintained manually on a regular basis. This can easily lead to failures being discovered untimely and failures being located inaccurately, leading to increased costs and disassembly risks caused by the spread or frequent inspections. Summary of the invention

[0005] In view of this, the present application provides a sensor optimization layout method and device for online status monitoring of a turbine cooler, which can monitor the operating status of the turbine cooler online, obtain fault information in real time, and achieve high-precision fault location.

[0006] Specifically, the present application is implemented through the following technical solutions:

[0007] A first aspect of the present application provides a method for optimizing sensor layout for online status monitoring of a turbine cooler, the method comprising:

[0008] Determine the sensor type based on the turbine cooler fault monitoring signal type and fault coverage index;

[0009] Establish a simulation model of the turbine cooler;

[0010] Acquire a fault signal, and input the fault signal into the simulation model, wherein the fault signal at least includes an operating parameter and a vibration parameter;

[0011] Identify fault-sensitive monitoring points based on correlation analysis between the output results of the simulation model and the fault signal, and determine sensor layout locations;

[0012] An online condition monitoring sensor for the turbine cooler is arranged based on the sensor layout position and the sensor type.

[0013] A second aspect of the present application provides a sensor optimization layout device for online status monitoring of a turbine cooler, the device comprising a determination module, an establishment module, an input module, an identification module and a layout module;

[0014] Wherein, the determination module is used to determine the sensor type based on the turbine cooler fault monitoring signal type and the fault coverage index;

[0015] The establishment module is used to establish a turbine cooler simulation model;

[0016] The input module is used to obtain a fault signal and input the fault signal into the simulation model, wherein the fault signal at least includes an operating parameter and a vibration parameter;

[0017] The identification module is used to identify fault-sensitive monitoring points based on the correlation analysis between the output result of the simulation model and the fault signal, and determine the sensor layout position;

[0018] The layout module is used to layout the online status monitoring sensor of the turbine cooler based on the sensor layout position and the sensor type.

[0019] The sensor optimization layout method and device for online status monitoring of turbine coolers provided in the present application can accurately collect fault signals and realize accurate fault judgment by selecting corresponding sensors that can sensitively identify the fault signal type of turbine coolers, and determine the sensor layout for health status identification from the fault mechanism analysis and the intelligent optimization algorithm analysis, which not only avoids the structural limitations, high economic costs and information redundancy caused by too many sensors installed, but also avoids the insufficient health status identification ability caused by too few sensors installed, and cannot realize early warning of faults, affecting system safety; in addition, the optimal measurement point is determined by considering multiple factors such as vibration transmission characteristics, fault mode and fault vibration characteristics, and actual installation conditions. First, the speed directly represents the operating condition of the turbine cooler. When the speed is abnormal, there must be something wrong with the product; secondly, the temperature signal. Since temperature monitoring has inertia and is easily affected by the environment, it can only be represented when the fault is serious to a certain extent. Therefore, it is necessary to directly collect signals at the fault heat source, otherwise it is easy to lose relevant information; vibration is more sensitive to faults and can be represented in the early stage of faults. However, the vibration signal is easily disturbed by the external working environment of the turbine and the transmission path, and the characteristics are not obvious. It is necessary to directly collect signals at the excitation source, and the typical fault of the turbine has a greater impact on radial vibration. Therefore, the fault can be identified only by collecting radial vibration, which can save the layout of axial sensors (first, the axial layout is difficult, and second, the signal characteristics are not obvious). In addition, by optimizing the sensor layout position, the sensitivity and accuracy of fault detection can be improved, false alarms and missed alarms can be reduced, and high-quality signal acquisition can be achieved, which is of great significance for directly obtaining signal fault characteristics, reducing internal and external environmental interference, supporting signal analysis, and achieving high-precision fault location. In this way, a sensor layout optimization method for online status monitoring of turbine coolers is established, which has high accuracy in health status recognition and can be used for real-time health status monitoring and fault diagnosis of turbine coolers during airborne use, supporting rapid fault location and maintenance guarantee of turbine coolers. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 A flow chart of Embodiment 1 of the method for optimizing sensor layout for online status monitoring of a turbine cooler provided in the present application;

[0021] Figure 2 The installation structure and measurement principle diagram of the rotation speed sensor shown in an exemplary embodiment of the present application;

[0022] Figure 3 This is a schematic diagram of the installation structure of a temperature sensor and a vibration sensor shown in an exemplary embodiment of the present application;

[0023] Figure 4 This is a structural schematic diagram of Example 1 of the sensor optimization layout device for online status monitoring of a turbine cooler provided in the present application. DETAILED DESCRIPTION

[0024] Here, exemplary embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application.

[0025] The terms used in this application are only for the purpose of describing specific embodiments and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in this article refers to and includes any or all possible combinations of one or more associated listed items.

[0026] It should be understood that although the terms first, second, third, etc. may be used in the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0027] Specific embodiments are given below to introduce the technical solution of the present application in detail.

[0028] Figure 1 This is a flow chart of the first embodiment of the sensor optimization layout method for online status monitoring of a turbine cooler provided in this application. Figure 1 , the method provided in this embodiment may include:

[0029] S101. Determine a sensor type based on a turbine cooler fault monitoring signal type and a fault coverage index.

[0030] It should be noted that the sensor type is determined based on the turbine cooler fault monitoring signal type and fault coverage index, including:

[0031] (1) Acquiring operating parameters of a turbine cooler and determining a fault monitoring signal type of the turbine cooler according to the sensitivity of the operating parameters to faults during operation; the fault monitoring signal type includes at least a speed signal and a vibration signal.

[0032] It should be noted that the turbine cooler is a core component of the aircraft environmental control system and is a typical high-speed rotating machine. Its typical failure modes are rotor imbalance, bearing wear, blade wear, etc. Specifically, the conventional operating parameters of the turbine cooler can be obtained by using various sensors and data acquisition systems, such as temperature sensors, pressure sensors, flow meters, etc., where the conventional operating parameters may include the inlet and outlet temperature, flow rate and pressure of the turbine end and the inlet and outlet temperature, flow rate and pressure of the fan end. However, these conventional parameters are not sensitive enough to the typical failures of the turbine cooler, and are prone to missed reports in the early stages of failures, resulting in untimely failure reporting and affecting mission safety.

[0033] Among them, sensitivity is to extract key features that can indicate the occurrence of faults, such as vibration amplitude, frequency domain feature changes, speed fluctuations, etc. after performing data preprocessing operations including denoising and normalization on the collected data. Through statistical analysis, machine learning or other algorithm analysis, the response degree of each parameter to different faults is evaluated to determine which parameters can best reflect specific faults.

[0034] Specifically, after analyzing and evaluating the response degree (i.e., sensitivity) of each parameter to different faults, the operating parameters can be sorted based on the sensitivity score, and the parameters with the highest sensitivity to specific faults can be preferentially selected as the fault monitoring signal type; at the same time, it should also be considered that the monitoring signal type can cover a variety of fault modes, and the signal is relatively stable and easy to measure. In this way, a comprehensive evaluation is performed on the most sensitive parameters to determine the final monitoring signal type (for example, in specific implementation, each condition amplitude can be considered, and the final monitoring signal type can be calculated by weighting).

[0035] For example, in one embodiment, the final monitoring signal type is determined to be a speed signal and a vibration signal according to the sensitivity. Therefore, in addition to the above-mentioned conventional operating parameters, speed signals and vibration signals that are more sensitive to typical faults of the turbine cooler can also be used as turbine cooler fault monitoring signals. Among them, the speed is one of the core parameters for measuring the working state of the turbine cooler, which can be used for both the control of the working state and the identification of the health state in the event of a fault such as a jam or overspeed. In addition, when a turbine cooler fails, the main feature is that the machine is accompanied by abnormal vibration and noise, and its vibration signal reflects the product's fault information from the amplitude domain, frequency domain and time domain. Vibration monitoring can be used to timely discover and identify these abnormal vibration phenomena, which is very important for evaluating the product's operating status and diagnosing faults. And because the turbine cooler has the characteristics of aerodynamic influence and large fluctuations in the working speed, the vibration signal will be affected by the external vibration environment and the transmission path of its own installation position, resulting in severe signal attenuation, resulting in unclear early fault characteristics. Therefore, the vibration of the fault source is measured first, and the bearing seat position is selected for vibration monitoring, which can monitor the rotor imbalance and bearing wear faults of the turbine cooler to the greatest extent. It should be noted that there is a certain correlation between the vibration signal and the rotational speed. By conducting in-depth analysis of the vibration signal, such as using a variety of signal processing techniques such as spectrum analysis and time domain analysis, the characteristics related to the periodic motion of the rotating parts can be extracted, and the rotational speed signal can be indirectly inferred. However, this process involves complex signal processing algorithms and professional signal processing knowledge, including signal acquisition, preprocessing, feature extraction, and rotational speed calculation. In contrast, directly using a speed sensor (such as photoelectric, magnetoelectric, and other speed sensors) for measurement, data acquisition is more direct, the operation is relatively simple, and in most cases the accuracy can meet engineering requirements, so direct measurement is generally used. Optionally, when the bearing of the turbine cooler fails due to wear, the internal temperature of the bearing will rise, so the bearing temperature can be used as one of the key monitoring signals for turbine cooler failure.

[0036] (2) Determine a fault coverage index of the turbine cooler according to the structure and limiting factors of the turbine cooler, wherein the limiting factors at least include reliability and cost factors.

[0037] It should be noted that the fault coverage index of the turbine cooler is determined according to the structure and limiting factors of the turbine cooler, wherein the structural factors may specifically include whether the sensor size and the installation position space match, and the fault coverage index refers to the proportion of fault types and fault areas that can be detected by the monitoring system, ensuring that the monitoring system has sufficient coverage capabilities for key faults, and the limiting factors include at least reliability and cost factors. Theoretically, the more sensors there are, the more system performance parameters can be obtained, and the more accurate and comprehensive the description of the health status of the system will be. However, for large and complex systems such as the PHM system of aerospace equipment, there are tens of thousands of monitoring variables. Due to limitations such as structural characteristics and economic cost factors, it is impossible to install sensors at all locations for measurement. In addition, the installation and layout of sensors may affect the working state of the equipment.

[0038] Specifically, it is necessary to understand the structure of the turbine cooler, including key components such as cooling channels, heat exchange areas, and turbine blades, and then identify potential faults based on structural characteristics, list possible fault types, such as thermal failure, abnormal vibration, leakage, wear, etc., and finally identify the key parts that are most sensitive to faults, such as turbine blades, cooling channel inlets and outlets, heat exchange surfaces, etc. In addition, it is also necessary to consider limiting factors, such as reliability and cost factors, including the total budget of the monitoring system (including sensor procurement, installation, maintenance, etc.). Specifically, when it is proved that the absence of a certain sensor is very effective for critical faults, reliability should be given priority, that is, whether the corresponding type of sensor can be installed, the weight and volume of the sensor, and the reliability environmental test performance. Only on this basis will cost factors be further considered. Specifically, as long as the sensor is useful for detecting critical faults, the benefits brought by its installation are far greater than the cost of the sensor itself; for example, other limiting factors include technical feasibility, data processing capabilities, system reliability (ensuring stable operation during long-term operation), and environmental adaptability of sensors.

[0039] In specific implementation, the internal structure of the turbine cooler can be understood in detail by obtaining the design drawings or CAD models of the turbine cooler, and the possible fault types can be determined based on the structural information. Then, the probability and severity of each fault type can be evaluated based on historical data or expert experience, and these faults can be ranked according to the degree of impact. A higher fault coverage target can be set for critical faults with a significant impact, and a lower fault coverage target can be set for faults with a smaller impact. A benefit analysis can be conducted in combination with the fault coverage target and limiting factors (such as economic cost), and the rationality of the set fault coverage target can be evaluated to optimize the coverage indicator.

[0040] (3) Determining a target sensor type based on the fault monitoring signal type and the fault coverage index.

[0041] It should be noted that determining the target sensor type based on the fault monitoring signal type and the fault coverage index specifically includes: establishing a relationship matrix between the turbine cooler fault mode and all sensors; determining the optimization target of the sensor based on the fault coverage index; and determining the target sensor type through a group optimization algorithm in combination with the optimization target.

[0042] Specifically, for example, in one embodiment, the fault modes can be determined by priority, including abnormal temperature, abnormal pressure, abnormal vibration, and abnormal rotation speed, and then the available sensor types that match the corresponding faults are selected, such as temperature sensors, pressure sensors, vibration sensors, and rotation speed sensors. Then, a relationship matrix between the fault mode and the sensor type is constructed, and the elements in the matrix represent the sensitivity of a certain sensor to a certain fault mode. For example, in one embodiment, the rows of the matrix represent the fault mode, and the columns represent the sensor type. The values ​​of the matrix elements can be 0 (no sensitivity), 1 (low sensitivity), 2 (medium sensitivity), 3 (high sensitivity), etc.

[0043] Based on the determined fault coverage index and the combined relationship matrix, it can be known which fault modes have not been effectively covered or the coverage is not ideal. Combined with actual needs and constraints (such as economic cost and technical feasibility), the optimization target of coverage is defined. For example, ensure that more than 90% of critical faults are effectively monitored. At this time, the optimization target is concretized to clarify the content to be optimized under the current constraints, such as improving the monitoring coverage of specific critical faults and reducing system costs; in addition, the priority of the optimization target can be determined to clarify which targets must be achieved and which targets can be optimized when conditions permit. Preferably, multiple evaluation indicators of fault isolation rate and fault diagnosis rate can also be combined to establish multiple optimization targets with the minimum number of sensor configurations, the lowest sensor weight, and the highest sensor reliability.

[0044] By using the improved bionic evolutionary swarm optimization algorithm, through chaos initialization, adaptive strategy to adjust inertia weight, mutation global search strategy, individual optimal and global optimal selection strategy under multi-objective game, multi-objective and multi-constrained sensor configuration optimization is performed to obtain the target sensor type that meets the fault diagnosis requirements. Specifically, these algorithms can find the optimal solution based on multiple constraints in a multi-dimensional search space, which is suitable for multi-objective optimization problems such as sensor type selection. In specific implementation, each individual in the group can be initialized, where each individual represents a sensor configuration scheme. For example, in one embodiment, each individual can be encoded as a vector, and each element of the vector represents whether a sensor type is selected. Then, an initial position is assigned to each individual, and an initial speed or other related parameters are set. The objective function is defined by combining factors such as fault coverage index, cost, and system complexity, and the sensor configuration scheme of each individual is brought into the objective function to calculate its fitness. Then, based on the swarm optimization algorithm, the individual position and speed are iteratively updated to continuously search for a better sensor configuration solution. In each iteration, the individual position (sensor configuration) is adjusted according to the fitness evaluation results, gradually approaching the optimal configuration (in the iterative process, the balance between the global optimum and the local optimum is considered to avoid falling into the local optimal solution). After a certain number of iterations or reaching the preset stop condition, the algorithm converges and finally outputs the global optimal sensor configuration solution. The final target sensor type is determined based on the optimization results.

[0045] In this way, under the condition of meeting the requirements of turbine cooler fault coverage index and other requirements, combined with the needs of fault monitoring signal type, selecting the appropriate target sensor type can achieve accurate fault monitoring, which has very important practical value.

[0046] S102. Establish a turbine cooler simulation model.

[0047] It should be noted that establishing a turbine cooler simulation model includes: analyzing the operating conditions of the various parts of the turbine cooler in combination with the stress and strain parameters and dynamic simulation of the various parts of the turbine cooler under the action of multi-field coupling; the operating conditions at least include the speed, acceleration and force conditions of the parts; analyzing the environmental factors of the turbine cooler in combination with the operating conditions; the environmental factors include environmental conditions, working loads and dynamic parameters; establishing a turbine cooler simulation model based on the multi-field coupling, dynamic simulation and environmental factors of the turbine cooler.

[0048] Specifically, in actual use, the operating environment of the turbine cooler is affected by loads such as high-speed airflow, ambient temperature and ambient vibration. Therefore, the establishment of a turbine cooler simulation model requires comprehensive consideration of multi-field coupling, dynamic simulation and environmental factors. In specific implementation, it is first necessary to study the stress and strain parameters of each part of the turbine cooler under the action of multi-field coupling. Secondly, it is necessary to further carry out multi-body dynamic simulation, analyze the speed, acceleration and force of the key parts in the turbine cooler, and obtain the operating conditions of each part of the turbine cooler. Finally, combined with the operating conditions, analyze the comprehensive influence of environmental conditions, working loads and dynamic parameters on the turbine cooler, and then establish a turbine cooler simulation model. Among them, the key parts include turbine blades, cooling channels, heat exchange areas, turbine casings, etc.

[0049] In specific implementation, the process of establishing the simulation model is as follows: establishing a geometric model of the turbine cooler for multi-field coupling analysis, including thermal-structural coupling simulation and fluid-structural coupling simulation; establishing a multi-body dynamics model of the turbine cooler for dynamic simulation analysis, including multi-body dynamics simulation (imposing dynamic loads such as rotational speed, gravity, aerodynamic force, etc. on the model, calculating the speed, acceleration and dynamic force of the parts) and vibration analysis; conducting environmental factor analysis, including operation under different ambient temperatures and pressures, working loads under different working conditions, changes in dynamic parameters under different operating modes, etc.; finally, integrating the multi-field coupling and dynamic simulation results, establishing a complete turbine cooler simulation model, integrating the thermal, structural, fluid, dynamic and other field results into one model.

[0050] S103, acquiring a fault signal, and inputting the fault signal into the simulation model, wherein the fault signal at least includes an operating parameter and a vibration parameter.

[0051] It should be noted that the fault signals of the turbine cooler under different fault conditions can be determined through historical data or expert experience. For example, common fault types and their corresponding operating parameters (including speed, temperature, pressure, flow, etc.) and vibration parameters (including vibration frequency, amplitude, acceleration, mode, etc.) can be identified. On this basis, the fault characteristics are extracted to obtain characteristic parameters related to various faults, such as blade fracture, bearing failure, thermal fatigue cracks, etc., and then the occurrence mechanism of each fault state is analyzed to determine its key characteristics and main influencing factors. For example, blade fracture is mainly caused by fatigue, overload, etc., and is characterized by local stress concentration and material fatigue failure; bearing failure is mainly caused by insufficient lubrication, overload, wear, etc., and is characterized by increased vibration and abnormal temperature increase; thermal fatigue cracks are caused by repeated thermal cycles, and are characterized by large local temperature gradients and stress concentration. Specifically, based on the analysis of typical key failure modes and mechanisms of turbine coolers, the occurrence process of failures can be simulated in the simulation model, and the input method in the simulation model can be clarified. For example, blade breakage can be achieved by reducing the material properties of the blades or introducing cracks in the geometric model; bearing failure can be achieved by increasing the vibration acceleration or frequency at the bearing and adjusting the friction coefficient of the bearing; thermal fatigue cracks can be achieved by applying periodic heat loads in local areas. In addition, leakage failures can be achieved by introducing fluid leakage paths or increasing channel flow resistance, and wear failures can be achieved by reducing the material hardness or thickness of the contact area. Specifically, the input methods of failures in the simulation model are not listed here one by one.

[0052] As an optional embodiment, the correlation analysis between the output result of the simulation model and the fault signal identifies fault-sensitive monitoring points and determines the sensor layout position, including: determining the real-time state of the turbine cooler according to the fault signal and its corresponding fault type, the fault signal characterizing the environmental conditions causing the fault; correcting the turbine cooler simulation model based on the real-time state, the real-time state being the equipment state of the turbine cooler before the fault corresponding to the fault type occurs; inputting the fault signal into the corrected turbine cooler simulation model, constraining the propagation path of the fault signal based on the fault signal propagation characteristics, and obtaining an output result, the output result at least including the state change of the corrected turbine cooler simulation model based on the fault signal; extracting multiple position point signals whose state change amplitude exceeds a threshold from the output result; determining the hierarchical relationship of the multiple position point signals based on the propagation characteristics of the fault signal in the turbine cooler; and determining the position point signal where the fault first occurs based on the hierarchical relationship as the fault-sensitive monitoring point. Specifically, when the fault occurs, the turbine cooler has been used for a period of time. Its state is not a brand new workpiece, but a worn one, and the device performance is degraded. At this time, in order to ensure that the output results of the simulation model are consistent with the evolution of the actual fault, the simulation model is first corrected according to the necessary equipment state generated by the fault type, and the simulation model is adjusted to the equipment state before the fault occurs. Secondly, after the condition of the fault, that is, the fault signal, is input into the simulation model, the fault-sensitive monitoring point is further determined according to the output result. Specifically, after the fault signal is input into the simulation model, a fault will occur inside the simulation model and propagate, that is, the input of the fault signal causes the propagation of the fault in the simulation model, which in turn causes inconsistent changes in the signal conditions at each position. On this basis, the position points with large changes are further screened out to monitor more serious faults. Among the position points with large changes, some positions are initial fault points, and some positions are secondary fault points caused by the initial fault points. At this time, the hierarchical relationship between multiple position points is analyzed according to the propagation distance of the fault, and the position point of the first layer is located as the initial fault point, so that the initial fault point is monitored as a fault-sensitive monitoring point, and the fault can occur at the earliest time. The method provided by the present invention does not require destructive observation or testing of the equipment. It can analyze and find the location point where the earliest fault occurs in the internal structure only through prior simulation and the assistance of historical fault data, thereby optimizing the layout of sensors and improving the efficiency, accuracy and timeliness of fault monitoring.

[0053] Specifically, combined with the established turbine cooler simulation model, the fault signals are input in sequence to ensure that only one fault is introduced each time, so as to facilitate the analysis of its individual impact. Then the simulation model is run to extract the operating parameters of the inlet and outlet interfaces under each fault state, such as pressure, temperature, flow, etc., and the vibration parameters under each fault state, including vibration spectrum, amplitude, frequency, etc. Among them, by comparing the operating parameters under the fault state and the normal state, the impact of the fault on the performance of the turbine cooler can be analyzed, and by comparing the vibration parameters under the fault state and the normal state, the fault characteristic frequency and amplitude changes can be identified.

[0054] Optionally, the characteristic parameters under various fault conditions can be organized into a fault characteristic library to facilitate fault diagnosis and prediction. Based on the fault characteristic library, a fault diagnosis algorithm is developed to monitor the operating status of the turbine cooler in real time and predict potential faults.

[0055] S104: Identify fault-sensitive monitoring points based on the correlation analysis between the output result of the simulation model and the fault signal, and determine the sensor layout position.

[0056] It should be noted that the identification of fault-sensitive monitoring points based on the correlation analysis between the output result of the simulation model and the fault signal and the determination of the sensor layout position include:

[0057] (1) Analyze the propagation mode and propagation path of the output result to obtain the analysis result.

[0058] The analysis of the propagation mode and propagation path of the output result to obtain the analysis result includes: obtaining the type of each fault signal in the output result; analyzing the propagation mode of different types of fault signals based on different media; establishing propagation models of various fault signals in different media, and obtaining their propagation modes and propagation paths based on the propagation models. It should be noted that the analysis process can be to determine the types of common fault signals, including vibration signals, temperature signals, pressure signals, etc., and then extract the characteristics of each signal, such as frequency range, amplitude, phase, etc., and then analyze the propagation mode of the signal based on different media (such as metal, gas, liquid, etc.), including reflection, refraction, attenuation and other phenomena, establish propagation models of various signals in different media, and analyze their propagation paths and propagation characteristics. On this basis, a fault source is set in the simulation model to simulate the generation and propagation of the fault signal, and by defining monitoring points at different positions, including various interfaces and surfaces of the turbine cooler, the propagation simulation of the fault signal is performed to obtain the signal response at each monitoring point.

[0059] (2) Determine the type and characteristics of the output result based on the analysis result.

[0060] It should be noted that the type of fault signal (vibration, temperature, pressure, etc.) and its characteristics (frequency, amplitude, phase, etc.) are extracted from the analysis results, and then the strength changes of the fault signal at different monitoring points are studied to identify the attenuation and enhancement characteristics of the signal during the propagation process.

[0061] (3) Obtaining data features of the fault signal and comparing them with the types and features contained in the output results to identify fault-sensitive monitoring points.

[0062] Specifically, by comparing the data characteristics of the fault signal with the simulation data under the fault state, and according to the propagation path of the fault characteristic signal and the change of signal strength, the monitoring points sensitive to the fault are identified.

[0063] (4) Determine the sensor layout location based on the distribution of the fault-sensitive monitoring points.

[0064] According to the distribution of fault-sensitive monitoring points, the optimal layout position of sensors is determined to ensure that key fault signals can be effectively captured and monitored.

[0065] In addition, it should be noted that the identification of fault-sensitive monitoring points based on the correlation analysis between the output result of the simulation model and the fault signal and the determination of the sensor layout position also include:

[0066] (1) Measuring actual fault monitoring parameters based on the determined sensor layout positions, and performing signal feature analysis with the output results of the simulation model.

[0067] Specifically, measuring actual fault monitoring parameters based on the determined sensor layout position includes:

[0068] A magnetoelectric speed sensor is respectively installed on the turbine disk partition of the turbine cooler, a temperature sensor (two sensors may be installed) is installed at two bearing positions inside the turbine cooler, and a vibration acceleration sensor is installed on the contact bearing intermediate housing of the turbine cooler; the pulse of the speed measuring rivet on the turbine disk is measured based on the magnetoelectric speed sensor, and the speed signal is calculated based on the pulse, the temperature change of the bearing is monitored based on the temperature sensor, and the vibration signal is monitored based on the vibration acceleration sensor.

[0069] Figure 2 This is an exemplary embodiment of the present application showing the installation structure and measurement principle diagram of the speed sensor. Figure 2As shown in Figure A and Figure B, a magnetoelectric speed sensor is installed on the turbine disk partition of the product. The product speed is detected by measuring the pulse of the speed measuring rivet on the turbine disk. Specifically, two steel rivets are symmetrically riveted on the product turbine disk. The speed sensor is installed on the partition, facing the speed measuring rivet of the turbine disk. When the turbine disk rotates, the sensor senses the change in magnetic flux when the speed measuring rivet passes by, forming a frequency pulse related to the speed, which is transmitted to the acquisition module for speed calculation.

[0070] Figure 3 This is a schematic diagram of the installation structure of a temperature sensor and a vibration sensor shown in an exemplary embodiment of the present application. Figure 3 Two PT1000 temperature sensors are installed at the two bearing positions inside the product to detect the temperature of the two bearings. The temperature sensor is inserted into the pin hole in the bearing housing, which is close to the bearing position. When the bearing wears out, the bearing temperature will continue to rise. The temperature measured at this location can directly indicate the product status. The inlet and outlet temperatures of the turbine cooler are mainly affected by the environment and the actual working performance of the product. They do not change much when the product is running, and cannot directly indicate whether the bearing is worn. Therefore, the optimal temperature measurement point is the bearing temperature.

[0071] Please continue to refer to Figure 3 , the vibration of the product's rotor is transmitted to the product's bearing housing through the product's bearing, and then to the intermediate housing. Therefore, the vibration acceleration sensor is installed on the bearing housing to measure the radial vibration of the product, reducing the intermediate transmission process, and more sensitively detecting the rotor vibration. When the product has critical faults such as imbalance and bearing wear, the source of the fault is directly transmitted to the bearing, and then directly transmitted to the bearing seat as the rigid body. The vibration here is the most direct and mainly concentrated in the radial vibration direction. If the vibration sensor is arranged on the turbine housing, the vibration from the excitation source (shaft, bearing) will undergo multiple stages of sensing and the signal will be severely attenuated. In addition, the turbine housing is easily affected by the internal airflow, resulting in inaccurate collected vibration signals and difficulty in identifying turbine faults.

[0072] (2) Adjusting the sensor layout position based on the analysis results.

[0073] It should be noted that, based on the results of the signal feature analysis, if new key monitoring points are identified, the sensor layout will be optimized. For example, if the signal features of certain parts have a low match with the fault state, additional sensors need to be installed in these parts. For another example, if the vibration features of a certain part fail to effectively capture the fault signal, additional sensors can be added to this part or the specific position and angle of the sensor can be adjusted.

[0074] Specifically, the sensor layout adjustment process can be as follows: based on the actual fault signal feature analysis results, evaluate the coverage capability and effectiveness of the current sensor layout; add new sensors or adjust the positions of existing sensors in areas where fault signals cannot be effectively captured to ensure that sufficient fault information can be obtained at key locations; under the adjusted layout, perform data collection and signal feature analysis again to confirm whether the adjustment has effectively improved the fault detection capability; through multiple iterations of adjustment and testing, gradually optimize the sensor layout to achieve the best fault detection coverage and accuracy.

[0075] S105. Arrange the online status monitoring sensor of the turbine cooler based on the sensor layout position and the sensor type.

[0076] It should be noted that the best sensor layout location (combined with the previous analysis results to determine which locations can most effectively monitor key faults, and select key components in the turbine cooler as the layout location of the sensor) and type are first determined through analysis, and then installed according to the detailed layout plan to ensure that each sensor can accurately collect the required fault signals. Secondly, the online monitoring system collects and analyzes data in real time to ensure that the health status of the turbine cooler is monitored and maintained in a timely manner.

[0077] The sensor layout optimization method for online status monitoring of a turbine cooler provided in this embodiment determines the sensor type based on the fault monitoring signal type and fault coverage index of the turbine cooler, establishes a turbine cooler simulation model, and then obtains the fault signal, inputs the fault signal into the simulation model, and then identifies the fault sensitive monitoring point based on the correlation analysis between the output result of the simulation model and the fault signal, determines the sensor layout position, and finally arranges the online status monitoring sensor of the turbine cooler based on the sensor layout position and the sensor type. In this way, by selecting the corresponding sensor that can sensitively identify the fault signal type of the turbine cooler, the fault signal can be accurately collected to achieve accurate fault judgment, and the sensor layout for health status identification is determined from the fault mechanism analysis and based on the intelligent optimization algorithm analysis, which not only avoids the structural limitations, high economic costs and information redundancy caused by too many sensors installed, but also avoids the insufficient health status identification ability caused by too few sensors installed, and the inability to achieve early warning of faults, affecting system safety; in addition, the optimal measurement point is determined by considering multiple factors such as vibration transmission characteristics, fault mode and fault vibration characteristics, and actual installation conditions. First, the speed directly represents the operating condition of the turbine cooler. When the speed is abnormal, there must be something wrong with the product. Secondly, the temperature signal. Since temperature monitoring has inertia and is easily affected by the environment, it can only be represented when the fault is serious to a certain extent. Therefore, it is necessary to directly collect signals at the fault heat source, otherwise it is easy to lose relevant information. Vibration is more sensitive to faults and can be represented in the early stage of the fault. However, the vibration signal is easily disturbed by the external working environment and transmission path of the turbine, and the characteristics are not obvious. It is necessary to collect signals directly at the excitation source. In addition, the typical fault of the turbine has a greater impact on radial vibration. Therefore, the fault can be identified only by collecting radial vibration, which can save the arrangement of axial sensors (first, the axial arrangement is difficult, and second, the signal characteristics are not obvious). In addition, by optimizing the sensor layout position, the sensitivity and accuracy of fault detection can be improved, false alarms and missed alarms can be reduced, and high-quality signal acquisition can be achieved, which is of great significance for directly obtaining signal fault characteristics, reducing internal and external environmental interference, supporting signal analysis, and achieving high-precision fault location. In this way, a sensor layout optimization method for online status monitoring of turbine coolers is established, which has high accuracy in health status identification and can be used for real-time monitoring of the health status and fault diagnosis of turbine coolers during airborne use, supporting rapid fault location and maintenance assurance of turbine coolers.

[0078] Corresponding to the aforementioned embodiment of a sensor optimization layout method for online status monitoring of a turbine cooler, the present application also provides an embodiment of a sensor optimization layout device for online status monitoring of a turbine cooler.

[0079] Figure 4This is a schematic diagram of the structure of the first embodiment of the sensor optimization layout device for online status monitoring of the turbine cooler provided by this application. Figure 4 , the device provided in this embodiment includes a determination module 410, an establishment module 420, an input module 430, an identification module 440 and a deployment module 450;

[0080] Wherein, the determination module 410 is used to determine the sensor type based on the turbine cooler fault monitoring signal type and the fault coverage index;

[0081] The establishment module 420 is used to establish a turbine cooler simulation model;

[0082] The input module 430 is used to obtain a fault signal and input the fault signal into the simulation model, wherein the fault signal at least includes an operating parameter and a vibration parameter;

[0083] The identification module 440 is used to identify fault-sensitive monitoring points based on the correlation analysis between the output result of the simulation model and the fault signal, and determine the sensor layout position;

[0084] The layout module 450 is used to layout the online status monitoring sensor of the turbine cooler based on the sensor layout position and the sensor type.

[0085] The device of this embodiment can be used to perform Figure 1 The steps, specific implementation principles and implementation processes of the method embodiment shown are similar and will not be repeated here.

[0086] The implementation process of the functions and effects of each unit in the above-mentioned device is specifically described in the implementation process of the corresponding steps in the above-mentioned method, and will not be repeated here.

[0087] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present application scheme. A person of ordinary skill in the art can understand and implement it without paying any creative work.

[0088] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for optimizing sensor layout for online status monitoring of a turbine cooler, characterized in that: The method comprises: Determine the sensor type based on the turbine cooler fault monitoring signal type and fault coverage index; Establish a simulation model of the turbine cooler; Acquire a fault signal, and input the fault signal into the simulation model, wherein the fault signal at least includes an operating parameter and a vibration parameter; Identify fault-sensitive monitoring points based on correlation analysis between the output results of the simulation model and the fault signal, and determine sensor layout locations; Arrange an online condition monitoring sensor for the turbine cooler based on the sensor layout position and the sensor type; The identifying of fault-sensitive monitoring points based on the correlation analysis between the output result of the simulation model and the fault signal and determining the sensor layout position include: determining a real-time state of the turbine cooler according to the fault signal and its corresponding fault type, the fault signal representing an environmental condition in which the fault occurs; Modifying the turbine cooler simulation model based on the real-time status, wherein the real-time status is the device status of the turbine cooler before the fault corresponding to the fault type occurs; Inputting the fault signal into the modified turbine cooler simulation model, constraining the propagation path of the fault signal based on the fault signal propagation characteristics, and obtaining an output result, wherein the output result at least includes a state change of the modified turbine cooler simulation model based on the fault signal; Extracting multiple position point signals whose state change amplitude exceeds a threshold from the output results; Determining the hierarchical relationship of the plurality of position point signals based on the propagation characteristics of the fault signal in the turbine cooler; Based on the hierarchical relationship, a location point signal where the fault occurs first is determined as a fault-sensitive monitoring point.

2. The method according to claim 1, characterized in that The method of determining the sensor type based on the turbine cooler fault monitoring signal type and the fault coverage index includes: Acquiring operating parameters of the turbine cooler and determining the type of fault monitoring signal of the turbine cooler according to the sensitivity of the operating parameters to faults during operation; the type of fault monitoring signal at least includes a speed signal and a vibration signal; Determining a fault coverage index of the turbine cooler according to the structure and limiting factors of the turbine cooler, wherein the limiting factors at least include reliability and cost factors; A target sensor type is determined based on the fault monitoring signal type and the fault coverage index.

3. The method according to claim 2, characterized in that The determining of the target sensor type based on the fault monitoring signal type and the fault coverage index specifically includes: Establish the relationship matrix between turbine cooler failure mode and all sensors; Determining an optimization target of the sensor based on the fault coverage index; In combination with the optimization objective, the target sensor type is determined by a group optimization algorithm.

4. The method according to claim 1, characterized in that The identifying of fault-sensitive monitoring points based on the correlation analysis between the output result of the simulation model and the fault signal and determining the sensor layout position include: Analyze the propagation mode and propagation path of the output result to obtain the analysis result; Determine the type and characteristics of the output result based on the analysis result; Acquire the data features of the fault signal and compare them with the types and features included in the output result to identify fault-sensitive monitoring points; The sensor layout positions are determined according to the distribution of the fault-sensitive monitoring points.

5. The method according to claim 4, characterized in that The analyzing the propagation mode and propagation path of the output result to obtain the analysis result includes: Get the type of each fault signal in the output result; Analyze the propagation modes of different types of fault signals based on different media; A propagation model of various fault signals in different media is established, and the propagation mode and propagation path thereof are obtained based on the propagation model.

6. The method according to claim 1, characterized in that The identifying of fault-sensitive monitoring points based on the correlation analysis between the output result of the simulation model and the fault signal and determining the sensor layout position also includes: Measuring actual fault monitoring parameters based on the determined sensor layout positions, and performing signal feature analysis with output results of the simulation model; The sensor layout position is adjusted based on the analysis result.

7. The method according to claim 1, characterized in that The step of establishing a turbine cooler simulation model comprises: Combined with the stress and strain parameters and dynamic simulation of each part of the turbine cooler under multi-field coupling, the operation conditions of each part of the turbine cooler are analyzed; the operation conditions at least include the speed, acceleration and force conditions of the parts; Analyzing environmental factors of the turbine cooler in combination with the operating conditions; the environmental factors include environmental conditions, workload and dynamic parameters; A turbine cooler simulation model is established based on the multi-field coupling, dynamic simulation and environmental factors of the turbine cooler.

8. The method according to claim 6, characterized in that The measuring of actual fault monitoring parameters based on the determined sensor layout position includes: A magnetoelectric speed sensor is installed on the turbine disk partition of the turbine cooler, a temperature sensor is installed at two bearing positions inside the turbine cooler, and a vibration acceleration sensor is installed on the contact bearing intermediate housing of the turbine cooler; The pulse of the speed measuring rivet on the turbine disk is measured based on the magnetoelectric speed sensor, and the speed signal is calculated according to the pulse. The temperature change of the bearing is monitored based on the temperature sensor. The vibration signal is monitored based on the vibration acceleration sensor.

9. A sensor optimization layout device for online status monitoring of a turbine cooler, characterized in that: The device includes a determination module, a creation module, an input module, an identification module and a deployment module; wherein, The determination module is used to determine the sensor type based on the turbine cooler fault monitoring signal type and the fault coverage index; The establishment module is used to establish a turbine cooler simulation model; The input module is used to obtain a fault signal and input the fault signal into the simulation model, wherein the fault signal at least includes an operating parameter and a vibration parameter; The identification module is used to identify fault-sensitive monitoring points based on the correlation analysis between the output result of the simulation model and the fault signal, and determine the sensor layout position; The layout module is used to layout the online status monitoring sensor of the turbine cooler based on the sensor layout position and the sensor type; The identifying of fault-sensitive monitoring points based on the correlation analysis between the output result of the simulation model and the fault signal and determining the sensor layout position include: determining a real-time state of the turbine cooler according to the fault signal and its corresponding fault type, the fault signal representing an environmental condition in which the fault occurs; Modifying the turbine cooler simulation model based on the real-time status, wherein the real-time status is the device status of the turbine cooler before the fault corresponding to the fault type occurs; Inputting the fault signal into the modified turbine cooler simulation model, constraining the propagation path of the fault signal based on the fault signal propagation characteristics, and obtaining an output result, wherein the output result at least includes a state change of the modified turbine cooler simulation model based on the fault signal; Extracting multiple position point signals whose state change amplitude exceeds a threshold from the output results; Determining the hierarchical relationship of the plurality of position point signals based on the propagation characteristics of the fault signal in the turbine cooler; Based on the hierarchical relationship, a location point signal where the fault occurs first is determined as a fault-sensitive monitoring point.

Images