CN-121981946-A - Underwater engine performance monitoring method based on wake feature recognition

Abstract

The invention discloses an underwater engine performance monitoring method based on wake feature recognition, which comprises the steps of S1, synchronously acquiring a thrust signal and a wake bubble evolution image sequence, S2, sequentially carrying out edge detection on the wake bubble evolution image sequence, extracting the outline boundary of a bubble to obtain wake bubble morphological feature parameters, constructing a wake bubble morphological feature sequence which changes along with time, S3, respectively carrying out transformation processing on the thrust signal and the wake bubble morphological feature sequence to obtain the frequency spectrum distribution features of the thrust signal and the wake bubble morphological feature sequence, S4, comparing the consistency of the thrust signal and the wake bubble morphological feature on the time domain trend and the frequency spectrum feature to form a mapping relation between the wake bubble morphological feature and the thrust signal, and S5, under the condition that the thrust signal cannot be acquired, carrying out speculation engine thrust by utilizing the mapping relation through the wake bubble image.

Inventors

• WANG JIE
• YIN CHAO
• CHEN DIAN
• CUI WEI
• WANG TIEHAN

Assignees

• 上海新力动力设备研究所

Dates

Publication Date

20260505

Application Date

20251212

Claims (10)

1. 1. An underwater engine performance monitoring method based on wake feature recognition is characterized by comprising the following steps: s1, acquiring a thrust signal of the engine in the whole underwater working process, simultaneously imaging a jet region of the engine in real time, and acquiring a wake bubble evolution image sequence formed under the jet effect; S2, sequentially carrying out edge detection on the wake flow bubble evolution image sequence, extracting the outline boundary of the bubble to obtain wake flow bubble morphological feature parameters, and constructing a wake flow bubble morphological feature sequence which changes along with time; s3, respectively carrying out transformation processing on the thrust signal and the wake flow bubble morphological feature sequence to obtain the spectrum distribution feature of the thrust signal and the spectrum distribution feature of the wake flow bubble morphological feature sequence; s4, comparing consistency of the thrust signal and the wake bubble morphological characteristics on time domain trend and frequency spectrum characteristics to form a mapping relation between the wake bubble morphological characteristics and the thrust signal; S5, under the condition that a thrust signal cannot be acquired, estimating the engine thrust by utilizing the mapping relation through the wake air bubble image.
2. 2. The method for monitoring the performance of the underwater engine based on wake feature recognition is characterized in that in the step S1, a thrust signal is obtained through a thrust sensor installed on an underwater thrust test platform, the wake bubble evolution image sequence is obtained through a high-speed imaging system, and the high-speed imaging system and the thrust sensor are triggered through a unified time synchronization mechanism so as to ensure that the time sequence of the signal and the time sequence of the image in the acquisition process are consistent.
3. 3. The method for monitoring performance of an underwater engine based on wake feature recognition as set forth in claim 1 wherein said thrust signal covers engine ignition, steady state combustion and extinction phases.
4. 4. The method for monitoring performance of an underwater engine based on wake feature recognition as set forth in claim 1, wherein in the step S2, the wake bubble morphological feature parameter is a width, an area or an aspect ratio of wake bubbles.
5. 5. The method for monitoring performance of an underwater engine based on wake feature recognition as set forth in claim 1, wherein the step S2 further comprises performing graying processing and median filtering processing on the wake bubble evolution image sequence before performing edge detection to eliminate interference of noise points and stray light in the underwater environment.
6. 6. The method for monitoring performance of an underwater engine based on wake feature recognition as set forth in claim 1, wherein in said step S2, an SOBEL operator is used for edge detection.
7. 7. The method for monitoring performance of an underwater engine based on wake feature recognition as set forth in claim 1, wherein in the step S3, a Fourier transform is used to obtain a spectral distribution feature of a thrust signal and a spectral distribution feature of a wake bubble morphological feature sequence.
8. 8. The method for monitoring performance of an underwater engine based on wake feature recognition as set forth in claim 1, wherein in the step S4, a Pelson correlation analysis method is adopted to compare consistency of the thrust signal and wake bubble morphological features in terms of time domain trend and spectral characteristics.
9. 9. Computer program product, characterized in that it realizes the steps of the method according to any of claims 1 to 8 when it is executed by a processor.
10. 10. The use of the method for monitoring the performance of an underwater engine based on wake feature recognition according to any one of claims 1 to 8 for monitoring the thrust of an underwater solid rocket engine or other underwater jet power device capable of forming wake features.

Description

Underwater engine performance monitoring method based on wake feature recognition Technical Field The invention relates to the technical field of solid rocket engines, in particular to an underwater engine performance monitoring method based on wake characteristic identification. Background In the research field of underwater solid rocket engines or other jet propulsion systems, accurate acquisition of thrust signals is always an important basis for performance evaluation and safe operation of the propulsion system. However, due to the complex and varying underwater environment, thrust sensors are often faced with multiple disturbances during acquisition. Specifically, factors such as water disturbance, environmental noise, vibration of underwater structures and the like can cause a large amount of non-stationary noise to be introduced into the sensor signal, so that effective signal components are covered. In this case, the key dynamic characteristic information is extremely easy to interfere and even lost, and the credibility and usability of the thrust data are seriously affected. The existing underwater solid rocket engine thrust measurement technical scheme is mostly dependent on direct measurement of a sensor, and lacks of utilization of obvious visual external phenomena in the underwater jet process, so that adaptability and robustness in a complex underwater environment are insufficient, and actual requirements of high-precision thrust monitoring and performance monitoring are difficult to meet. Disclosure of Invention The invention solves the technical problems of overcoming the defects of the prior art, providing the underwater engine performance monitoring method based on wake characteristic identification, breaking through the limitation that a single sensor acquisition path is easy to be interfered by noise, realizing underwater thrust monitoring through multi-source information fusion, and improving the accuracy and the robustness of overall signal interpretation. The technical scheme of the invention is that the method for monitoring the performance of the underwater engine based on wake characteristic identification comprises the following steps: s1, acquiring a thrust signal of the engine in the whole underwater working process, simultaneously imaging a jet region of the engine in real time, and acquiring a wake bubble evolution image sequence formed under the jet effect; S2, sequentially carrying out edge detection on the wake flow bubble evolution image sequence, extracting the outline boundary of the bubble to obtain wake flow bubble morphological feature parameters, and constructing a wake flow bubble morphological feature sequence which changes along with time; s3, respectively carrying out transformation processing on the thrust signal and the wake flow bubble morphological feature sequence to obtain the spectrum distribution feature of the thrust signal and the spectrum distribution feature of the wake flow bubble morphological feature sequence; s4, comparing consistency of the thrust signal and the wake bubble morphological characteristics on time domain trend and frequency spectrum characteristics to form a mapping relation between the wake bubble morphological characteristics and the thrust signal; S5, under the condition that a thrust signal cannot be acquired, estimating the engine thrust by utilizing the mapping relation through the wake air bubble image. In step S1, a thrust signal is obtained by installing a thrust sensor on an underwater thrust test platform, the wake bubble evolution image sequence is obtained by a high-speed imaging system, and the high-speed imaging system and the thrust sensor are triggered by a uniform time synchronization mechanism to ensure that the signal is consistent with the time sequence of the image in the acquisition process. Further, the thrust signal covers engine ignition, steady state combustion, and extinction phases. Further, in the step S2, the wake bubble morphological feature parameter is a width, an area or an aspect ratio of the wake bubble. Further, in the step S2, before the edge detection, the method further includes performing graying processing and median filtering processing on the wake bubble evolution image sequence, so as to eliminate interference of noise points and stray light in the underwater environment. Further, in the step S2, an SOBEL operator is used to perform edge detection. Further, in the step S3, a fourier transform is adopted to obtain a spectrum distribution characteristic of the thrust signal and a spectrum distribution characteristic of the wake bubble morphological characteristic sequence. Further, in the step S4, a pearson correlation analysis method is adopted to compare the consistency of the thrust signal and the wake bubble morphology feature in the time domain trend and the frequency spectrum characteristic. The invention also relates to a computer program product which, when being executed by a proc