CN-122014428-A - Isolation section intelligent suction control method for shock wave string position feedback

CN122014428ACN 122014428 ACN122014428 ACN 122014428ACN-122014428-A

Abstract

The invention relates to the technical field of active flow control of supersonic isolation segments, and discloses an intelligent pumping control method of an isolation segment for shock wave string position feedback, which comprises the steps of acquiring wall pressure data of the isolation segment by using synchronous data acquisition equipment, and transmitting the wall pressure data to an edge computing platform for preprocessing; the method comprises the steps of inputting data into a lightweight flow field reconstruction model to generate a flow field schlieren image, utilizing a target detection algorithm to identify wave system characteristics, calculating the front edge position of a shock wave string through a confidence weighting algorithm, generating a feedback control instruction according to the relative relation between the front edge position of the shock wave string and the position of a suction slot, and driving an electromagnetic valve to open and close through a serial port relay to realize intelligent suction regulation. The invention realizes accurate capture and real-time closed-loop control of the shock wave string position through reconstruction and weighted positioning from sparse pressure to dense flow field based on an edge computing architecture, effectively inhibits the shock wave string from jumping upstream, and maintains the stability of the flow field of the isolation section.

Inventors

KONG CHEN
CHANG JUNTAO
WANG ZIAO
LV CHENGKUN

Assignees

哈尔滨工业大学

Dates

Publication Date: 20260512
Application Date: 20260224

Claims (10)

1. The intelligent pumping control method for the isolation section of the shock wave string position feedback is characterized by comprising the following steps of: Acquiring wall pressure data output by the wall pressure sensor (10) of the isolation section in real time by using synchronous data acquisition equipment (20), and transmitting the wall pressure data to an edge computing platform (30) for preprocessing; Inputting the preprocessed wall pressure data into a lightweight flow field reconstruction model deployed on the edge computing platform (30), and outputting flow field schlieren images reflecting the density gradient change in the isolation section; Identifying the wave system characteristics in the flow field schlieren image by using a target detection algorithm, and calculating the position of the front edge of the shock wave string by using a confidence weighting algorithm; generating a feedback control instruction according to the relative relation between the front edge position of the shock wave string and a preset pumping groove flow direction position interval; And the feedback control instruction is issued by the serial port relay (40) to drive the electromagnetic valve (50) to open and close, so that intelligent suction regulation and control of the flow field of the isolation section are realized.
2. The intelligent pumping control method for the isolated segment with shock wave string position feedback according to claim 1, wherein the process of acquiring wall pressure data output by the isolated segment wall pressure sensor (10) in real time by using the synchronous data acquisition equipment (20) and transmitting the wall pressure data to the edge computing platform (30) for preprocessing comprises the following steps: The synchronous data acquisition equipment (20) carries out analog-to-digital conversion on the analog voltage signal output by the wall surface pressure sensor (10) according to a preset sampling frequency, encapsulates the converted digital pressure data and sends the digital pressure data to the edge computing platform (30) through a TCP/IP protocol; The edge computing platform (30) establishes a first-in first-out buffer pool for the data channel of each wall surface pressure sensor (10), executes a moving average filtering algorithm on the received original pressure data sequence, and calculates an arithmetic average value in a filtering window as effective pressure data.
3. The method of intelligent pumping control of an isolated segment for shock string position feedback according to claim 2, wherein the process of preprocessing by the edge computing platform (30) further comprises: Dividing the effective pressure value of each wall pressure sensor (10) subjected to moving average filtering by the real-time inlet static pressure data at the same moment by using the real-time inlet static pressure data acquired by the wall pressure sensor (10) arranged at the inlet of the isolation section as a reference quantity to obtain a dimensionless wall pressure ratio; and linearly mapping the dimensionless wall pressure ratio to a value interval from zero to one by using a maximum and minimum normalization algorithm, and generating a normalized pressure characteristic vector as an input of the lightweight flow field reconstruction model.
4. The intelligent pumping control method for the isolated segment with shock wave string position feedback according to claim 1, wherein the process of inputting the preprocessed wall pressure data into a lightweight flow field reconstruction model deployed on the edge computing platform (30) and outputting flow field schlieren images reflecting density gradient changes in the isolated segment comprises the following steps: inputting the wall pressure data into a convolutional neural network branch of the lightweight flow field reconstruction model, performing up-sampling operation through a full-connection layer dimension lifting and transposition convolutional layer, and outputting two-dimensional space structural characteristics of a flow field; inputting the wall pressure data into a fully-connected neural network branch of the lightweight flow field reconstruction model, and extracting global nonlinear correlation characteristics between the wall pressure data through a plurality of fully-connected layers; And splicing the two-dimensional space structural features and the global nonlinear correlation features at a fusion layer of the lightweight flow field reconstruction model, and decoding through an output convolution layer to generate the flow field schlieren image.
5. The isolated segment intelligent pumping control method of shock string position feedback of claim 4, wherein the lightweight flow field reconstruction model performs a model weight process prior to deployment on the edge computing platform (30), the model weight process comprising: Analyzing the weight sparsity of a model convolution layer, identifying and eliminating redundant characteristic channels with contribution degree lower than a preset threshold value to an output result, and reducing floating point operation times of forward reasoning of the model; And (3) semi-precision quantization, namely converting weight parameters and bias parameters in the model from a 32-bit floating point number format to a 16-bit floating point number format, compiling the converted data into a file format suitable for mobile terminal deployment, and reducing the memory throughput delay of the edge computing platform (30).
6. The method of claim 1, wherein the identifying the wave system features in the flow field schlieren image by using the target detection algorithm comprises: running a target detection algorithm based on YOLOv architecture, and executing double-category target recognition on the flow field schlieren image; Identifying a first detection category, and marking the first detection category as a background wave system area, wherein the background wave system area is an upstream area which is not influenced by the main disturbance of the shock wave string in the isolation section or a conventional background wave system structure area; identifying a second detection category, and marking the second detection category as a shock wave string dominant region, wherein the shock wave string dominant region is a region comprising a core shock wave structure and an induced boundary layer separation region; and outputting the prediction boundary frame coordinates of each region and the corresponding classification confidence value.
7. The method of claim 6, wherein the calculating the front edge position of the shock string by using a confidence weighting algorithm comprises: extracting the right edge coordinate of the prediction boundary frame of the background wave system region on the flow direction coordinate axis and the classification confidence corresponding to the background wave system region; Extracting left edge coordinates of a prediction boundary box of the shock wave string dominant region on a flow direction coordinate axis and classification confidence corresponding to the shock wave string dominant region; and constructing a weighted average correction model, and carrying out weighted calculation on the right edge coordinate and the left edge coordinate by using the classification confidence coefficient as a weight factor to obtain the corrected shock string front edge position.
8. The intelligent pumping control method of the isolation section for the shock wave string position feedback according to claim 1, wherein the process of generating the feedback control command according to the relative relation between the shock wave string front edge position and the preset pumping groove flow direction position interval comprises the following steps: Comparing the position of the front edge of the shock wave string with the flow direction position interval of the suction groove; When the front edge position of the shock wave string is smaller than or equal to the flow direction position interval of the suction groove, judging that the flow field of the isolation section is in an anti-back pressure demand state, generating an opening instruction and setting the position of a control mark as 1; when the front edge position of the shock wave string is larger than the flow direction position interval of the suction groove, judging that the flow field of the isolation section is in a safe area state, generating a closing instruction and setting the position of a control mark as 0.
9. The intelligent pumping control method of the isolated segment of the shock wave string position feedback according to claim 8, wherein the process of sending the feedback control command to drive the electromagnetic valve (50) to open and close through the serial port relay (40) comprises the following steps: the edge computing platform (30) sends hexadecimal control messages to the serial port relay (40) through a ModbusRTU protocol; when the control zone bit is 1, the serial port relay (40) responds to the opening instruction to drive the internal physical contact to be closed, a power supply loop of the electromagnetic valve (50) is connected, and a pneumatic pipeline channel connected to the isolation section suction groove is opened; when the control flag bit is 0, the serial port relay (40) responds to the closing instruction to drive the internal physical contact to be disconnected, the power supply loop of the electromagnetic valve (50) is cut off, and the electromagnetic valve (50) is reset to be closed.
10. The method for intelligent pumping control of an isolated segment for shock wave string position feedback according to claim 1, wherein the arrangement mode of the wall pressure sensor (10) comprises the following steps: Arranging the wall pressure sensor (10) at the inlet of the isolation section, and acquiring incoming flow static pressure data as a reference quantity; Selecting a plurality of measuring points which are distributed at equal intervals on the lower wall surface of the isolation section to arrange the wall surface pressure sensor (10) so as to acquire flow direction pressure distribution data in the isolation section; Arranging the wall pressure sensor (10) at the outlet of the isolation section, and monitoring outlet backpressure; the wall pressure sensors (10) at the locations described above together form a sparse pressure input array and are physically connected to the synchronous data acquisition device (20) by signal cables.

Description

Isolation section intelligent suction control method for shock wave string position feedback Technical Field The invention relates to the technical field of active flow control of supersonic isolation segments, in particular to an intelligent pumping control method of an isolation segment for shock wave string position feedback. Background At present, in the operation process of the air suction hypersonic propulsion system, the isolation section is used as a key pneumatic component for connecting the air inlet channel and the combustion chamber and bears the important tasks of accommodating the shock wave string and isolating the combustion back pressure. The flow direction position of the shock wave string structure in the isolation section is not fixed, but dynamically moves along with the fluctuation of the flight Mach number and the back pressure of the combustion chamber. Maintaining a stable position of the shock wave string in the isolation section, preventing excessive forward movement from causing the inlet channel to be not started, is a core element for ensuring the safe operation of the propulsion system. To address the risk of shock string migration, it is often necessary to implement flow control on the separator wall to enhance the back pressure resistance of the flow field by removing the low energy boundary layer. Aiming at the monitoring and control requirements of the flow field of the isolation section, the prior technical proposal mostly adopts a detection mode based on a discrete wall pressure sensor to cooperate with passive or simple open loop active control. In practice, the wall pressure distribution is obtained by arranging an array of pressure sensors in the flow direction. The specific pressure threshold is set by a pressure jump method or a standard deviation method. When the pressure at a certain measuring point exceeds a threshold value, the front edge of the shock wave string is judged to reach the position. Based on the judging result, part of the schemes adopt a passive suction chamber with fixed aperture ratio to continuously suck, or simply control the opening and closing of the valve according to a preset Mach number working condition table, and the real-time analysis of the details of the convection field is not involved. However, the above prior art has limitations in facing highly dynamically changing complex flow fields. The discrete pressure measuring points have sparsity in space, and the complete flow field topology containing shock wave/boundary layer interference details is difficult to reconstruct, so that dead zones exist in cognition of the state of the flow field. The discrimination logic which only depends on the pressure threshold value is sensitive to signal noise, and the fuzzy physical boundary of the front edge of the shock wave string cannot be accurately defined, so that the severe jump and false alarm of the positioning result are easy to cause. In addition, control schemes lacking real-time perception of flow field morphology tend to respond with hysteresis. While passive pumping is simple, it can cause continuous flow loss and reduce thrust performance of engine, simple switch control is difficult to accurately anchor the shock wave string position, and when encountering sudden high back pressure, it is unable to effectively restrain various unsteady oscillations and sudden jump movements of the shock wave string through fast and accurate flow field feedback. Therefore, the invention provides an intelligent pumping control method for the isolation section of the shock wave string position feedback, which solves the defects in the prior art. Disclosure of Invention Aiming at the defects of the prior art, the invention provides an intelligent suction control method for an isolation section of shock wave string position feedback, which solves the problems that the shock wave string positioning precision is low and the shock wave string is difficult to effectively inhibit the shock wave string from jumping upstream in real time to cause the non-starting of an air inlet channel due to the sparse discrete pressure monitoring information in a supersonic speed isolation section. The invention aims at realizing the technical scheme that the intelligent pumping control method for the isolation section of the shock wave string position feedback comprises the following steps: Acquiring wall pressure data output by a wall pressure sensor of an isolation section in real time by using synchronous data acquisition equipment, and transmitting the wall pressure data to an edge computing platform for preprocessing; inputting the preprocessed wall pressure data into a lightweight flow field reconstruction model deployed on the edge computing platform, and outputting a flow field schlieren image reflecting the density gradient change in the isolation section; Identifying the wave system characteristics in the flow field schlieren image by using a target detection alg