A method and system for extracting characteristic points of electromagnetic valve current curve

Abstract

This invention discloses a method and system for extracting characteristic points from a solenoid valve current curve. The disclosed scheme obtains a characteristic signal extraction curve by filtering the solenoid valve current curve, performing first-order differentiation, three-layer wavelet packet decomposition, signal reconstruction, and absolute value processing. The system then uses a sliding window and the accumulated voltage within the window to locate the pull-in and release points within the characteristic signal extraction curve. This method achieves high accuracy and is applicable to a variety of solenoid valve types.

Description

Technical Field

[0001] The invention relates to a method for identifying characteristic points of a current curve of a solenoid valve of a liquid rocket engine. Background Art

[0002] The electromagnet in a solenoid valve is a coil with an iron core, which has a certain inductance. When voltage is applied to the coil, the current in the coil increases from zero, and the magnetic flux increases accordingly. This increase in magnetic flux generates an induced potential in the coil, which prevents the current from increasing. Therefore, the current cannot reach a stable value directly, but instead gradually increases to a stable value according to the law of an exponential curve.

[0003] As the current increases, the suction force on the armature gradually increases. When the suction force is greater than the reaction force, the armature begins to move. Since the movement of the armature causes the inductance to change, a back electromotive force is generated. This back electromotive force prevents the current from rising. Therefore, when the armature starts to move, the current not only does not continue to rise but gradually decreases. After the armature is fully attracted, the current continues to rise. This turning point is called the attraction point.

[0004] The faster the armature moves, the greater the back EMF generated and the faster the current drops;

[0005] When the coil is de-energized, the current gradually decreases from a stable value, and the magnetic flux also gradually decreases from a stable magnetic flux. When the magnetic flux decreases to a point where the suction force it generates cannot hold the armature, the armature begins to release under the action of the spring force. During the release process, a back electromotive force will still be generated. At this time, the back electromotive force prevents the current from decreasing, so the current will rebound until the armature returns to its original position and the current continues to drop to 0. This turning point is called the release point.

[0006] The current curve of a solenoid valve during its engagement or release should typically have only one peak. However, because the peaks in the solenoid valve current curve are caused by the movement of the valve core, multiple peaks may appear during mass production. If the valve continuously accelerates under the action of a reaction force, it will exhibit a single peak (i.e., a single-peak valve). If the valve does not continuously accelerate and undergoes multiple decelerations, it will exhibit multiple peaks (i.e., a multi-peak valve).

[0007] To determine the reliability and performance of a solenoid valve, it's necessary to extract characteristic points, such as the pull-in and release points. Existing methods for interpreting the pull-in and release characteristics of current curves use a method that works backwards from the moment the solenoid valve starts actuating. The difference between the values ​​of the subsequent and previous points is greater than a certain threshold. When the difference between the values ​​of the subsequent and previous points is less than the threshold, the inflection point peak is considered to have been found. The difference between the values ​​of the subsequent and previous points is less than a certain threshold. When the difference between the values ​​of the subsequent and previous points is greater than the threshold, the inflection point trough is considered to have been found. The time corresponding to the trough is subtracted from the value at the start time to determine the solenoid valve opening time. The closing time is the reverse process of the opening time method.

[0008] The above-mentioned existing methods place high demands on the filtering of the current curve. The curve must be monotonically increasing or decreasing, otherwise misjudgment may occur, resulting in a high misjudgment rate. Due to the different characteristics of different valve models, the threshold must be continuously adjusted to obtain the appropriate threshold for different valve types in different engine models, increasing the difficulty of adjustment and the complexity of model management. Furthermore, if the solenoid valve is switched from a single-peak valve to a multi-peak valve, the entire recognition algorithm must be replaced. Summary of the Invention

[0009] In view of the defects or shortcomings of the prior art, the present invention provides a method for extracting characteristic points of a solenoid valve current curve, which is used to extract the pull-in point and the release point of the solenoid valve current curve.

[0010] To this end, the method provided by the present invention comprises:

[0011] Step 1: Perform sliding average filtering on the voltage value of the solenoid valve current curve;

[0012] Step 2, performing first-order differentiation on the filtered data in step 1 to obtain a current curve change rate curve;

[0013] Step 3, decompose the current curve change rate curve into 3-layer wavelet packets to obtain 8 wavelet nodes;

[0014] Step 4, reconstruct the signals of the 8 wavelet nodes obtained in step 3 respectively to obtain the signal waveforms of the 8 nodes;

[0015] Step 5: Perform absolute value processing on the second node signal waveform among the 8 node signal waveforms to obtain the characteristic signal extraction curve Y data ;

[0016] Step 6, extracting the pull-in point, including:

[0017] Step 61: Using WinWidth sampling points as a sliding window, slide from left to right on the characteristic signal extraction curve starting from the beginning of the curve, sliding one sampling point at a time. Each time the sliding window is opened, the cumulative sum sum1 of each voltage value in the current sliding window on the characteristic signal extraction curve is calculated. WinWidth takes 0.4%-0.6% of the total number of sampling points; Y data (i) is the voltage value of sampling point i on the characteristic signal extraction curve in the current sliding window;

[0018] Step 62: When the cumulative sum of the voltage values ​​in the current sliding window is greater than or equal to 5max(Y data ), extract the left edge point of the current sliding window LeftEdge and stop sliding; max(Y data ) is the maximum voltage value on the characteristic signal extraction curve; otherwise, continue sliding until the left boundary point LeftEdge is found;

[0019] Step 63, extracting the characteristic signal curve Y data Extract the horizontal coordinate PointPullX_L corresponding to the maximum voltage value between the point LeftEdge and the point WinWidth behind the point LeftEdge;

[0020] Step 64 , mapping PointPullX_L to the solenoid valve current curve, extracting the abscissa PointPullX corresponding to the minimum voltage value between the point PointReleX_L and the point WinWidth after PointReleX_L on the solenoid valve current curve, which is the abscissa of the pull-in point;

[0021] Step 7, extracting the release point, includes:

[0022] Step 71: Using WinWidth sampling points as a sliding window, slide from the end of the characteristic signal extraction curve from right to left, sliding one sampling point at a time. Each time the sliding window is opened, the cumulative sum sum2 of each voltage value in the current sliding window on the characteristic signal extraction curve is calculated. Y data (j) is the voltage value of sampling point j on the characteristic signal extraction curve in the current sliding window;

[0023] Step 72: When the cumulative sum sum2 in the current sliding window is greater than or equal to 5max(Y data ), extract the right boundary point RightEdge of the current sliding window and stop sliding; otherwise, continue sliding until the right boundary point RightEdge is found;

[0024] Step 73, extract the characteristic signal curve Y data Extract the horizontal coordinate PointReleX_R corresponding to the maximum voltage value between the WinWidth point before the point RightEdge and the point RightEdge;

[0025] Step 74 , map PointReleX_R to the solenoid valve current curve data, and extract the abscissa PointtReleX corresponding to the maximum voltage value between the point WinWidth before the point PointReleX_R on the solenoid valve current curve and the point PointReleX_R. This abscissa is the abscissa of the release point.

[0026] Furthermore, WinWidth is set to 0.5% of the total number of sampling points.

[0027] Furthermore, for a multi-peak valve, the above method is used to extract the first attraction point from left to right and the first release point from right to left on the current curve; then the sliding window continues to slide from left to right, and steps 61-64 are executed to find all the attraction points in sequence; the sliding window continues to slide from right to left, and steps 71-74 are executed to find all the release points in sequence.

[0028] The present invention also provides a system for extracting characteristic points from a solenoid valve current curve. The system includes a data preprocessing module, an engagement point extraction module, and a release point extraction module. The data preprocessing module is used to perform steps 1-5 above, the engagement point extraction module is used to perform step 6 above, and the release point extraction module is used to perform step 7 above.

[0029] Compared with the existing technology, the method of the present application has the following advantages: (1) By using wavelet packet decomposition and reconstruction, the solenoid valve current curve is decomposed into the time domain and frequency domain, thereby extracting the characteristic points of the solenoid valve's engagement and release, and improving the accuracy of characteristic point identification; (2) This method is applicable to various types of solenoid valves and can identify single-peak valves and multi-peak valves, thereby improving the adaptability of the identification method. (3) By using wavelet packets to decompose the current curve into various frequency domains for analysis, there is no need to extract characteristic points through thresholds, thus eliminating the difficulty of threshold adjustment. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a waveform diagram of the third layer signal after decomposition and reconstruction of the single-peak valve wavelet packet in Example 1 of the present invention;

[0031] Figure 1 (a) is the filtered curve of Example 1; Figure 1 (b) is the curve after first-order differentiation of Example 1; Figure 1 (c)-(j) are the signal waveforms of the eight nodes in Example 1 respectively;

[0032] Figure 2 This is a diagram showing the absolute value of the second node signal of the single-peak valve in Example 1 of the present invention;

[0033] Figure 3 This is a diagram showing the result of identifying characteristic points of a single-peak solenoid valve current curve according to Example 1 of the present invention;

[0034] Figure 4 This is a waveform diagram of the third layer signal after bimodal valve wavelet packet decomposition and reconstruction according to Example 2 of the present invention;

[0035] Figure 4 (a) is the filtered curve of Example 2; Figure 4 (b) is the curve after first-order differentiation of Example 2; Figure 4 (c)-(j) are the signal waveforms of the eight nodes in Example 2 respectively;

[0036] Figure 5 This is a diagram showing the absolute value of the second node signal of the bimodal valve according to embodiment 2 of the present invention;

[0037] Figure 6 This is a diagram showing the result of identifying characteristic points of the current curve of the double-peak valve solenoid valve according to Example 2 of the present invention. DETAILED DESCRIPTION

[0038] Unless otherwise specified, the terms used herein are understood according to the knowledge of a person skilled in the art.

[0039] The term "from left to right" used herein refers to the direction from the beginning to the end of the curve, and the term "from right to left" refers to the direction from the end to the beginning of the curve.

[0040] The current curve of the present invention is obtained by converting the collected voltage curve; the conversion refers to the conversion of the relationship between current and voltage. The three-layer wavelet packet decomposition, reconstruction, and absolute value processing of the present invention all adopt existing related methods.

[0041] Example:

[0042] This embodiment uses the method of the present invention to extract the pull-in point and release point of a single-peak solenoid valve used in a rocket engine. The steps are as follows:

[0043] Step 1: Perform sliding average filtering on the collected voltage value of the electromagnetic valve current curve, see Figure 1 (a);

[0044] Step 2: Perform first-order differentiation on the filtered data to obtain the current curve change rate curve, see Figure 1 (b);

[0045] Step 3, decompose the current curve change rate curve into 3-layer wavelet packets to obtain 8 wavelet nodes;

[0046] Step 4: reconstruct the signals of the 8 wavelet nodes obtained in step 3 respectively to obtain the signal waveforms of the 8 nodes, see Figure 1 (c)-(j);

[0047] Step 5, select the second node signal waveform ( Figure 1 (d) Perform absolute value processing to obtain the characteristic signal extraction curve Y data ,See Figure 2 ;

[0048] Then, the pull-in point and release point of the single-peak solenoid valve are extracted by using steps 6 and 7 of the present invention, wherein the number of WinWidth is 0.5% of the total number of sampling points; the result is as follows: Figure 3 shown.

[0049] Furthermore, the pull-in time and release time of the single-peak solenoid valve are calculated: pull-in time = the corresponding time of the pull-in point - the corresponding time of the power-on point; release time = the corresponding time of the power-off point - the corresponding time of the release point.

[0050] Example 2:

[0051] This embodiment uses the method of the present invention to extract the two pull-in points and two release points of the double-peak solenoid valve used in the rocket propulsion system. The specific steps are as follows:

[0052] Using the method described in steps 1-5, the 8 wavelet node signals of the electromagnetic valve to be tested and the absolute value processing of the second node signal are obtained. The results are shown in Figure 4 and Figure 5 ;

[0053] Then, use steps 6 and 7 to extract the first pull-in point from left to right and the first release point from right to left on the curve, where the number of WinWidth is 0.5% of the total number of sampling points;

[0054] Then continue to slide the sliding window from left to right, execute steps 61-64 to find the second suction point; continue to slide the sliding window from right to left, execute steps 71-74 to find the second release point, the result is shown in Figure 6 .

[0055] Furthermore, the pull-in time and release time of the bimodal solenoid valve are calculated:

[0056] Attracting time 1 = the corresponding time of the first attracting point - the corresponding time of the power-on point; releasing time 1 = the corresponding time of the power-off point - the corresponding time of the first releasing point.

[0057] Attracting time 2 = the corresponding moment of the second attracting point - the corresponding moment of the power-on point; releasing time 2 = the corresponding moment of the power-off point - the corresponding moment of the second releasing point.

Claims

1. A method for extracting characteristic points of a solenoid valve current curve, the method is used to extract the pull-in point and release point of the solenoid valve current curve, characterized in that: The method comprises: Step 1: Perform sliding average filtering on the voltage value of the solenoid valve current curve; Step 2, performing first-order differentiation on the filtered data in step 1 to obtain a current curve change rate curve; Step 3, decompose the current curve change rate curve into 3-layer wavelet packets to obtain 8 wavelet nodes; Step 4, reconstruct the signals of the 8 wavelet nodes obtained in step 3 respectively to obtain the signal waveforms of the 8 nodes; Step 5: Perform absolute value processing on the second node signal waveform among the 8 node signal waveforms to obtain the characteristic signal extraction curve Y data ; Step 6, extracting the pull-in point, including: Step 61: Using WinWidth sampling points as a sliding window, slide from left to right on the characteristic signal extraction curve starting from the beginning of the curve, sliding one sampling point at a time. Each time the sliding window is opened, the cumulative sum sum1 of each voltage value in the current sliding window on the characteristic signal extraction curve is calculated. WinWidth takes 0.4%-0.6% of the total number of sampling points; Y data (i) is the voltage value of sampling point i on the characteristic signal extraction curve in the current sliding window; Step 62: When the cumulative sum of the voltage values ​​in the current sliding window is greater than or equal to 5max(Y data ), extract the left edge point of the current sliding window LeftEdge and stop sliding; max(Y data ) is the maximum voltage value on the characteristic signal extraction curve; otherwise, continue sliding until the left boundary point LeftEdge is found; Step 63, extracting the characteristic signal curve Y data Extract the horizontal coordinate PointPullX_L corresponding to the maximum voltage value between the point LeftEdge and the point WinWidth behind the point LeftEdge; Step 64 , mapping PointPullX_L to the solenoid valve current curve, extracting the abscissa PointPullX corresponding to the minimum voltage value between the point PointReleX_L and the point WinWidth after PointReleX_L on the solenoid valve current curve, which is the abscissa of the pull-in point; Step 7, extracting the release point, includes: Step 71: Using WinWidth sampling points as a sliding window, slide from the end of the characteristic signal extraction curve from right to left, sliding one sampling point at a time. Each time the sliding window is opened, the cumulative sum sum2 of each voltage value in the current sliding window on the characteristic signal extraction curve is calculated. Y data (j) is the voltage value of sampling point j on the characteristic signal extraction curve in the current sliding window; Step 72: When the cumulative sum sum2 in the current sliding window is greater than or equal to 5max(Y data ), extract the right boundary point RightEdge of the current sliding window and stop sliding; otherwise, continue sliding until the right boundary point RightEdge is found; Step 73, extract the characteristic signal curve Y data Extract the horizontal coordinate PointReleX_R corresponding to the maximum voltage value between the WinWidth point before the point RightEdge and the point RightEdge; Step 74 , map PointReleX_R to the solenoid valve current curve data, and extract the abscissa PointtReleX corresponding to the maximum voltage value between the point WinWidth before the point PointReleX_R on the solenoid valve current curve and the point PointReleX_R. This abscissa is the abscissa of the release point.

2. The method for extracting characteristic points of a solenoid valve current curve according to claim 1, wherein: WinWidth is 0.5% of the total number of sampling points.

3. The method for extracting characteristic points of a solenoid valve current curve according to claim 1, wherein: For a multi-peak valve, the method described in claim 1 is used to extract the first attraction point from left to right and the first release point from right to left on the current curve; then the sliding window continues to slide from left to right, and steps 61-64 are executed to find all the attraction points in sequence; the sliding window continues to slide from right to left, and steps 71-74 are executed to find all the release points in sequence.

4. A solenoid valve current curve feature point extraction system, characterized in that: It includes a data preprocessing module, an attraction point extraction module and a release point extraction module; the data preprocessing module is used to execute steps 1-5 of claim 1, the attraction point extraction module is used to execute step 6 of claim 1, and the release point extraction module is used to execute step 7 of claim 1.

Images