Movable hot spot simulation experiment system and method for real-time sensing of turbine thermal field
The mobile hot spot simulation experimental system utilizes mechanical positioning and distributed sensors to achieve real-time sensing of the turbine thermal field of an aero-engine, solving the problems of flexibility and real-time performance in hot spot sensing in existing technologies and providing an efficient verification platform.
Patent Information
- Application Number
- CN202511324506.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-17
- Publication Date
- 2026-01-06
AI Technical Summary
Existing technologies struggle to accurately and in real-time detect the location and intensity of hot spots at the turbine inlet in aero-engines, and lack flexible experimental systems and automated real-time data links, resulting in high algorithm verification costs, high resource consumption, and large model errors.
A mobile hot spot simulation experimental system was designed, which includes a precisely locatable heat source, multi-point temperature detection, and a real-time data link based on file polling. An automated closed-loop verification platform was constructed, and the precise location of the hot spot and real-time data acquisition were achieved by using a mechanical positioning mechanism and distributed temperature sensors.
It achieves accurate simulation and real-time sensing of hot spot location, reduces experimental costs, improves system flexibility and automation, and ensures online performance testing and verification of the algorithm.
Smart Images
Figure CN121279079A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of engineering thermophysics and intelligent testing technology, and specifically relates to an experimental system for simulating the hot spot at the inlet of an aero-engine turbine and for real-time, closed-loop verification of the thermal field sensing algorithm. More specifically, it is a movable hot spot simulation experimental system for real-time sensing of the turbine thermal field. Background Technology
[0002] In the design and operation of advanced aero-engines, the temperature distribution (i.e., the thermal field) at the turbine inlet is extremely uneven, often exhibiting localized high-temperature regions, known as "hot spots." Hot spots can seriously threaten the structural safety and service life of turbine blades. Therefore, the ability to perceive the location and intensity of upstream hot spots in real time is crucial for engine health management and performance optimization.
[0003] In recent years, intelligent sensing algorithms (such as those based on artificial intelligence and Kalman filtering) that use limited downstream sensor data to invert or predict the complete upstream thermal field have been extensively studied. However, the development and validation of such algorithms face significant challenges:
[0004] Real-world experiments are costly: conducting experiments on real aircraft engines or gas turbines is extremely costly and risky, and the placement of sensors is limited, making it difficult to obtain high-quality verification data.
[0005] Pure numerical simulation has limitations: Although computational fluid dynamics (CFD) simulation can provide full-field data, it consumes huge computational resources, making it difficult to run in real time, and there is always a model error between its simulation results and physical reality.
[0006] Existing experimental setups lack flexibility: Traditional laboratory simulation devices typically have fixed structures, making it difficult to adjust the position of the heat source. This makes it difficult to easily simulate the occurrence of hot spots in different locations, resulting in incomplete coverage of algorithm verification conditions and difficulty in achieving automated testing.
[0007] The lack of a real-time data link: In algorithm verification, a major technical challenge is how to seamlessly, automatically, and with low latency transmit real-time sensor data generated by the physical experimental platform into the running sensing algorithm. Traditional methods often rely on an offline mode of "acquisition-stop-manual export-import algorithm-computation," which cannot truly assess the online real-time performance of the algorithm.
[0008] Therefore, there is an urgent need for a physical simulation experimental system that is simple in structure, cost-controllable, flexibly deployable, and equipped with an automated real-time data link, so as to provide an efficient and reliable verification platform for advanced thermal field sensing algorithms. Summary of the Invention
[0009] To overcome the shortcomings of the prior art, the present invention aims to provide a mobile hot spot simulation experimental system for real-time sensing of turbine thermal fields. The system integrates a precisely located heat source, distributed temperature measurement, and an innovative real-time data link based on file polling, thus constructing an automated closed-loop verification platform from physical simulation to real-time sensing.
[0010] To achieve the above objectives, the present invention adopts the following technical solution:
[0011] A movable hot spot simulation experimental system for real-time sensing of turbine thermal field is characterized by comprising: a hot spot generator, a multi-point temperature detection unit, and a processing unit; the hot spot generator, which can move and be precisely positioned within a two-dimensional plane, is disposed in the upstream region of the simulated turbine flow channel; the multi-point temperature detection unit includes temperature sensors for real-time acquisition of temperature distribution data of the downstream section, the temperature sensors being arranged in an array in the downstream region of the simulated turbine flow channel; the processing unit is connected to the data output terminal of the multi-point temperature detection unit and stores and processes the data output by the multi-point temperature detection unit;
[0012] The processing unit includes a real-time data acquisition program for continuously monitoring a preset data storage folder and automatically identifying and reading data files containing the latest temperature distribution data generated by the multi-point temperature detection unit from the folder; a thermal field sensing model for receiving temperature distribution data read by the real-time data acquisition program as input and calculating the predicted location of the hot spot generator upstream; and a real-time result display interface for visually displaying the predicted location calculated by the thermal field sensing model in real time.
[0013] Furthermore, the multi-point temperature detection unit includes 16 temperature sensors, which are arranged in an array in the downstream region of the simulated turbine flow channel to collect temperature distribution data of the downstream section in real time.
[0014] The hot spot generator is equipped with a heating unit. The hot spot generator is set in the upstream area of the simulated turbine flow channel through a mechanical positioning mechanism, and the two-dimensional movement and precise positioning of the hot spot generator are achieved through the cooperation of guide rails, sliders and locking nuts.
[0015] The data files generated by the multi-point temperature detection unit are named with filenames containing timestamps or serial numbers, so that the real-time data acquisition program can identify the latest data files.
[0016] The real-time data acquisition program determines the latest data file by comparing the filenames or modification times of data files within the folder.
[0017] The thermal field sensing model is a virtual-real fusion model that has been pre-trained with computational fluid dynamics (CFD) simulation data and fine-tuned with a small amount of collected experimental data.
[0018] The multi-point temperature detection unit continuously generates data files to the data storage folder. The processing unit captures the latest data files through the real-time data acquisition program, sends the acquired temperature distribution data into the thermal field sensing model for online calculation, and dynamically updates and displays the predicted upstream hot spot location through the real-time result display interface.
[0019] The hot spot generator is moved sequentially to a series of known physical locations (x). i ,y i It also records the downstream temperature distribution data T corresponding to each known physical location, measured by the multi-point temperature detection unit. i This forms a one-to-one corresponding training data pair ((x i ,y i ),T i ), generate training data.
[0020] The hot spot generator is fixed at a known physical location, and its physical location is compared with the predicted location output by the model on the real-time result display interface to verify the accuracy of the thermal field sensing model.
[0021] A method for a movable hot spot simulation experimental system for real-time sensing of turbine thermal fields, characterized by comprising the following steps:
[0022] Step 1: Positioning. Move the hot spot generator to a preset position upstream of the simulated turbine flow channel and fix it in place.
[0023] Step 2: Measurement and Recording. Start the system, and the multi-point temperature detection unit continuously collects downstream temperature distribution data. The collected data is continuously generated and saved in the form of data files to a preset data storage folder.
[0024] Step 3: Monitoring and reading. The data acquisition program in the processing unit monitors the data storage folder frequently. Once a new data file is detected, it is immediately read to obtain the latest temperature distribution data.
[0025] Step 4: Prediction. The temperature distribution data read in Step 3 is used as input and fed into the thermal field sensing model for calculation to obtain the predicted location of the upstream hot spot.
[0026] Step 5: Display. The predicted position obtained in Step 4 is dynamically updated and displayed on the real-time result display interface to achieve real-time perception.
[0027] In this invention, the processing unit is associated with the data output terminal of the multi-point temperature detection unit. The processing unit stores and runs: a real-time data acquisition program, a thermal field sensing model, and a real-time result display interface.
[0028] During system operation, the multi-point temperature detection unit continuously generates data files containing the latest temperature distribution data to a preset data storage folder; the processing unit continuously monitors the folder through a real-time data acquisition program in a high-frequency polling manner. Once a new data file is detected, it automatically reads it and sends the temperature data in it into the thermal field sensing model for online calculation; finally, the upstream hot spot location predicted by the model is dynamically visualized by the real-time result display interface.
[0029] Compared with the prior art, the beneficial effects of the present invention are as follows:
[0030] 1. Flexible structure and precise positioning: The mechanical positioning mechanism consisting of guide rails, sliders and locking nuts can easily and accurately move the hot spot source to any specified position, easily simulate various working conditions, and provide comprehensive test data for the algorithm.
[0031] 2. Highly automated and robust data link: It uniquely adopts a decoupled data interaction method of "continuous writing to files at the data acquisition end and high-frequency reading of files at the algorithm end", which avoids complex network programming and communication protocol configuration. The system has extremely high stability, is simple to deploy, and is very suitable for laboratory environments.
[0032] 3. Achieve true real-time closed-loop verification: It breaks down the "last mile" from physical signals to algorithm input, enabling perception algorithms to undergo true online performance testing and verification in a dynamically changing physical environment, rather than traditional offline analysis.
[0033] 4. Versatile functions and high cost performance: This system can not only be used to verify the accuracy and real-time performance of existing algorithms, but also to efficiently collect training data. By moving the heat source to a series of known locations and recording the corresponding downstream temperature fields, it can quickly generate high-quality "location-temperature" data pairs for training and optimizing new sensing models. Attached Figure Description
[0034] Figure 1 This is a schematic diagram of the overall system structure according to an embodiment of the present invention.
[0035] Figure 2 This is a schematic diagram of the mechanical structure of the movable hot spot generator in an embodiment of the present invention.
[0036] Figure 3This is a system workflow diagram of an embodiment of the present invention.
[0037] Figure 4 This is a schematic diagram of the software interface of the host computer processing unit in an embodiment of the present invention. Detailed Implementation
[0038] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.
[0039] Reference Figure 1 and Figure 2 This embodiment describes a movable hot spot simulation experimental system for real-time sensing of turbine thermal fields. The system is built on a simulated turbine flow channel 4 and mainly includes a hot spot generator 1, a multi-point temperature detection unit 2, and a processing unit 3.
[0040] A hot spot generator 1 is used to simulate a localized high-temperature hot spot upstream of the turbine. In this embodiment, the hot spot generator uses a high-power heating coil as its heating element 11. The hot spot generator 1 is positioned in the upstream region of the simulated turbine flow channel 4 via a mechanical positioning mechanism, and its two-dimensional movement and precise positioning are achieved through the cooperation of the guide rail 13, the slider 12, and the locking nut 14. To achieve movable and precise positioning, the hot spot generator 1 moves horizontally (X-direction) by adjusting the relative position of the horizontal guide rail and the slider; and moves vertically (Y-direction) by adjusting the pair of vertically adjustable locking nuts.
[0041] Precise scales are engraved next to the guide rail 13. The slider 12 can be moved manually, and after reaching the target position in the horizontal direction, it is positioned and fixed in the vertical direction by locking the nut 14. This purely mechanical positioning method is simple in structure, low in cost, and reliable in positioning. Figure 2 This is a schematic diagram of the mechanical structure of the movable hot spot generator in an embodiment of the present invention.
[0042] The multi-point temperature detection unit 2 is used to acquire the temperature distribution of the downstream thermal field. In this embodiment, it consists of 16 4×4 array K-type thermocouples as temperature sensors 21, which are uniformly arranged on the downstream cross-section of the simulated turbine flow channel 4. The signal lines of all temperature sensors 21 are connected to a data acquisition device 22. The data acquisition device 22 acquires the temperature values of all sensors at a set frequency and continuously writes the 16 temperature values at each moment as a data row to a CSV format data.csv file located on the local disk of the processing unit 3. Each write overwrites the old file to ensure that the file always contains the latest temperature data.
[0043] Processing unit 3 is a personal computer. Its core is a software program running on it, which includes three main modules: a real-time data acquisition program 31, a thermal field sensing model 32, and a real-time result display interface 33.
[0044] Reference Figure 3 The workflow of this system is as follows:
[0045] Step S101 (Locating the hot spot): Before the experiment begins, the operator manually moves the hot spot generator 1 to a known physical location (x) in the upstream flow channel. true ,y true And secure it with lock nut 14.
[0046] Step S102 (Start Data Acquisition): Turn on the hot spot generator 1 and the temperature data acquisition device 22. The temperature data acquisition device 22 starts to continuously measure the downstream temperature field at a certain frequency and writes the latest 16 temperature values in real time to the data.csv file in the designated shared folder.
[0047] Step S103 (Monitoring and Reading): Simultaneously, the real-time data acquisition program 31 on the processing unit 3 continuously checks the "last modified time" of the data.csv file at a certain frequency. Once a change in this timestamp is detected, the program immediately determines that new data has been written and quickly opens and reads the latest line of temperature data T from the file. new This file-based polling mechanism cleverly decouples the data acquisition hardware from the upper-layer application software.
[0048] Step S104 (Model Prediction): The real-time data acquisition program 31 reads the temperature data array T new As input, it is passed to the pre-loaded thermal field sensing model 32. This model is a neural network that has been trained using CFD simulation data and real experimental data. The model receives T new Then, a forward inference calculation is performed, outputting a predicted upstream hotspot location (x). pred ,y pred ).
[0049] Step S105 (Result Visualization): Refer to Figure 4 Processing unit 3 will predict the position (x) pred ,y pred The result is plotted in real-time on the real-time display interface 33, represented by a red dot. The interface can also simultaneously display a representation of the actual physical location (x-axis). true ,y true The image shows the predicted location. It allows for a direct comparison between the predicted and actual locations, and the distance between them can be calculated as the prediction error, thus evaluating the performance of the perception model.
[0050] Step S106 (loop): The system continuously repeats steps S103 to S105, thereby realizing continuous and real-time sensing and display of hot spot location.
[0051] In addition, this system can also be used to generate training data. The operator simply needs to move the hotspot generator 1 sequentially to a series of pre-planned physical location points (x... i ,y i After stabilization at each location point, the steady-state downstream temperature distribution data T measured by the multi-point temperature detection unit 2 is recorded. i In this way, a large number of one-to-one correspondences ((x) can be generated efficiently and at low cost. i ,y i ),T i Training data pairs are used to train and optimize the perception algorithm.
[0052] In summary, through ingenious mechanical design and innovative software architecture, this invention successfully constructs a fully functional, easy-to-operate, and robust mobile hot spot simulation experimental system, providing an important experimental platform for the development of turbine thermal field real-time sensing technology.
[0053] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A movable hot spot simulation experiment system for turbine hot field real-time perception, characterized in that, It comprises a hot spot generator (1), a multi-point temperature detection unit (2) and a processing unit (3); the hot spot generator (1) which can move in a two-dimensional plane and be accurately positioned is arranged in the upstream area of the simulated turbine flow passage (4); the multi-point temperature detection unit (2) comprises temperature sensors (21) for collecting temperature distribution data of the downstream section in real time, and the temperature sensors (21) are arranged in an array in the downstream area of the simulated turbine flow passage (4); the processing unit (3) is connected with the data output end of the multi-point temperature detection unit (2) and stores and runs the data output by the multi-point temperature detection unit (2); The processing unit (3) comprises a data real-time acquisition program (31) for continuously monitoring a preset data storage folder and automatically identifying and reading a data file generated by the multi-point temperature detection unit (2) and containing the latest temperature distribution data from the folder; a thermal field perception model (32) for receiving the temperature distribution data read by the data real-time acquisition program (31) as input and calculating the predicted position of the hot spot generator (1) in the upstream; A result real-time display interface (33) for visually displaying the predicted position calculated by the thermal field perception model (32) in real time. The hot spot generator (1) is provided with a heating unit (11), and the hot spot generator (1) is arranged in the upstream area of the simulated turbine flow passage (4) through a mechanical positioning mechanism and realizes two-dimensional movement and accurate positioning of the hot spot generator (1) through cooperation of a guide rail (13), a sliding block (12) and a locking nut (14).
2. The movable hot spot simulation experiment system for real-time perception of a turbine hot field according to claim 1, characterized in that, The data file generated by the multi-point temperature detection unit (2) is named with a file name with a time stamp or a serial number, so as to facilitate the data real-time acquisition program (31) to identify the latest data file.
3. The movable hot spot simulation experiment system for real-time perception of a turbine hot field according to claim 1, characterized in that, The data real-time acquisition program (31) determines the latest data file by comparing the file name or file modification time of the data file in the folder.
4. The movable hot spot simulation experiment system for real-time perception of a turbine hot field according to claim 3, characterized in that, The thermal field perception model (32) is a virtual-real fusion model which is trained in advance through computational fluid dynamics simulation data and fine-tuned in combination with a small amount of experimental data.
5. The movable hot spot simulation experiment system for real-time perception of a turbine hot field according to claim 1, characterized in that, The multi-point temperature detection unit (2) continuously generates data files to the data storage folder, the processing unit (3) captures the latest data file through the data real-time acquisition program (31), sends the acquired temperature distribution data to the thermal field perception model (32) for online calculation, and dynamically updates and displays the predicted upstream hot spot position on the result real-time display interface (33).
6. The movable hot spot analog test system for turbine hot gas path real-time sensing of claim 1, wherein, The hot spot generator (1) is fixed at a known physical position, and its physical position is compared with the predicted position output by the model on the result real-time display interface (33) to verify the accuracy of the thermal field perception model (32).
7. The movable hot spot simulation experiment system for real-time perception of a turbine hot field according to claim 1, characterized in that, The hot spot generator (1) is moved to a series of known physical positions (x i ,y i ) in turn, and the downstream temperature distribution data T i measured by the multi-point temperature detection unit (2) corresponding to each known physical position is recorded to form a one-to-one corresponding training data pair ((x i ,y i ), T i ), and training data is generated.
8. The movable hot spot simulation experiment system for real-time perception of a turbine hot field according to claim 1, characterized in that, It comprises the following steps:
9. A method for a movable hot spot simulation experiment system for real-time perception of a turbine hot field based on the system of claim 1, characterized in that, Step one: positioning, moving the hot spot generator to a preset position in the upstream of the simulated turbine flow passage and fixing it; Step two: measurement and record, start the system, continuously collect downstream temperature distribution data by the multi-point temperature detection unit, and continuously generate and save the collected data in the form of data files to a preset data storage folder; Step three: monitoring and reading, through the data real-time collection program in the processing unit, high-frequency monitoring of the data storage folder, once detecting the generation of a new data file, immediately reading the file to obtain the latest temperature distribution data; Step four: prediction, taking the temperature distribution data read in step three as input, sending it into the thermal field perception model for calculation to obtain the predicted position of the upstream hot spot; Step five: display, dynamically updating and displaying the predicted position obtained in step four on the result real-time display interface, thereby realizing real-time perception.