A method for identifying edge wear of a last-stage blade of a low-pressure cylinder of a steam turbine

Abstract

The method for identifying the edge wear of the last-stage blade of a low-pressure cylinder of a steam turbine comprises the following steps: obtaining picture information of the last-stage blade of the steam turbine and converting the picture information into a grayscale image; performing square filtering and gradient calculation on the grayscale image; obtaining an edge curve and a convex hull line of the blade and calculating a centroid; calculating a distance and angle sequence formed by the centroid and the edge curve and the convex hull line; obtaining a starting point sequence number of the edge curve and a terminal point sequence number of the edge curve by using filtering and extreme value screening; obtaining a first closed curve coordinate sequence and a second closed curve coordinate sequence composed of the edge curve and the convex hull line and the centroid; finally obtaining an edge wear area by using a Gaussian shoe formula, and determining whether the edge wear reaches a set alarm threshold value, and timely performing alarm; the method can quickly and accurately identify the edge wear of the last-stage blade of the low-pressure cylinder of the steam turbine, improve the identification efficiency, reduce the input of manpower and material resources, and has important significance for the intelligent construction of power plants.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of turbine low-pressure cylinder last-stage blade edge wear identification, and particularly relates to a turbine low-pressure cylinder last-stage blade edge wear method. BACKGROUND

[0002] The turbine low-pressure cylinder last-stage blade is an important core component of a turbine, has a relatively long length, and works in a humid environment. The working environment is relatively complex, and the working condition of the low-pressure last-stage area is more severe due to the demand for peak regulation in recent years. The main failure form of the turbine low-pressure cylinder last-stage blade is blade edge wear, and the main causes of the blade edge wear are blade water erosion and foreign object impact. Severe wear may even lead to blade fracture. Currently, some turbine manufacturers have installed endoscope devices on the last-stage blades to monitor and evaluate the edge wear state of the turbine in real time. However, the identification of the turbine low-pressure cylinder last-stage blade edge wear in the prior art still remains in the stage of manual identification. The manual identification method is too subjective, has low accuracy, and cannot quantify the wear amount. In addition, the manual identification method needs to invest a large amount of manpower and material resources, and does not meet the demand for the intelligence of power plants. SUMMARY

[0003] To solve the problem of low identification efficiency of the method for identifying the turbine low-pressure cylinder last-stage blade edge wear by manual identification in the prior art, a turbine low-pressure cylinder last-stage blade edge wear method is provided.

[0004] A turbine low-pressure cylinder last-stage blade edge wear identification method, comprising:

[0005] Step one: obtaining image information of a turbine last-stage blade, and converting the obtained image information into a gray image;

[0006] Step two: obtaining an image matrix R of the gray image obtained in step one; obtaining an edge curve according to the image matrix R, obtaining a convex hull line and a centroid according to the edge curve;

[0007] Step three: calculating a included angle and distance data according to the edge curve, the convex hull line and the centroid, wherein the included angle and distance data comprise an included angle sequence of the convex hull line and the centroid, an included angle sequence of the edge curve and the centroid, and a distance sequence of the edge curve and the centroid;

[0008] Step four: filtering and extreme value screening the included angle and distance data to obtain a sequence number of a blade edge starting point in the included angle sequence of the edge curve and the centroid and a sequence number of a blade edge ending point in the included angle sequence of the edge curve and the centroid;

[0009] Step 5: Based on the sequence numbers of the angle sequence between the blade edge starting point and the edge curve and the centroid, and the sequence number of the angle sequence between the blade edge ending point and the edge curve and the centroid, truncate the angle sequence between the edge curve and the centroid and the distance sequence between the convex hull and the centroid to obtain sequences L1 and L3 respectively; add the centroid after sequence L1 to obtain the first closed curve coordinate sequence; add the starting point, ending point, and centroid of sequence L1 after sequence L3 to obtain the second closed curve coordinate sequence;

[0010] Step 6: Calculate the first closed area using the Gaussian shoelace formula based on the coordinate sequence of the first closed curve; calculate the second closed area using the Gaussian shoelace formula based on the coordinate sequence of the second closed curve; and obtain the edge wear area by subtracting the first and second closed areas.

[0011] Step 7: Determine if the edge wear area is greater than the set threshold; if so, issue an alarm.

[0012] Beneficial effects

[0013] This invention acquires image information of the last-stage blades of a steam turbine and converts it into a one-dimensional grayscale image. Block filtering and gradient calculation are performed on the grayscale image to obtain an image matrix, aiming to eliminate noise and expose the edges of the metal blades. Erosion and dilation calculations are performed on the image matrix to close the metal edges of the blades. The edge curves and convex hull lines of the blades are calculated to determine the centroid. The distance and angle sequences formed by the centroid, edge curves, and convex hull lines are calculated. Filtering and extreme value sieving are used to obtain the starting point sequence number of the edge curve and the ending point sequence number of the edge curve. The coordinate sequences of the first and second closed curves formed by the edge curves, convex hull lines, and centroids are obtained. The area of ​​the closed curves is calculated using the Gaussian shoelace formula, ultimately obtaining the edge wear area. The method then determines whether the edge wear of the last-stage blades of the low-pressure cylinder of the steam turbine has reached a set alarm threshold and issues an alarm in a timely manner. This method can quickly and accurately identify the edge wear of the last-stage blades of the low-pressure cylinder of a steam turbine, improving identification efficiency and reducing the investment of manpower and resources, which is of great significance for the intelligent construction of power plants. Attached Figure Description

[0014] Figure 1 A flowchart illustrating a method for identifying edge wear of the last stage blades of a low-pressure cylinder in a steam turbine, as described in a specific embodiment of this application;

[0015] Figure 2 The image obtained by performing block filtering on a grayscale image according to a specific embodiment of this application;

[0016] Figure 3 The image obtained after gradient processing for a specific embodiment of this application;

[0017] Figure 4An image obtained after expansion for the specific embodiment of the present application;

[0018] Figure 5 An image of the position of the sequence number of the starting point of the edge curve and the sequence number of the ending point of the edge curve for the specific embodiment of the present application;

[0019] Figure 6 An image of the blade edge wear area for the specific embodiment of the present application;

[0020] Figure 7 An image of the blade edge non-wear area for the specific embodiment of the present application. DETAILED DESCRIPTION

[0021] The specific embodiment will be described below in conjunction with Figures 1 to 7 A turbine low-pressure cylinder last-stage blade edge wear identification method, comprising:

[0022] Step one: obtaining image information of the turbine last-stage blade, and converting the obtained image information into a gray-scale image;

[0023] Step two: obtaining an image matrix R of the gray-scale image obtained in step one; obtaining an edge curve according to the image matrix R, and obtaining a convex hull line and a centroid according to the edge curve;

[0024] Step three: calculating an included angle and distance data including a sequence of included angles of the convex hull line and the centroid, a sequence of included angles of the edge curve and the centroid, and a sequence of distances of the edge curve and the centroid according to the edge curve, the convex hull line and the centroid;

[0025] Step four: filtering and extreme value screening the included angle and distance data to obtain a sequence number of a blade edge starting point in the sequence of included angles of the edge curve and the centroid and a sequence number of a blade edge ending point in the sequence of included angles of the edge curve and the centroid;

[0026] Step five: intercepting the sequence of included angles of the edge curve and the centroid and the sequence of distances of the convex hull line and the centroid according to the obtained sequence number of the blade edge starting point in the sequence of included angles of the edge curve and the centroid and the sequence number of the blade edge ending point in the sequence of included angles of the edge curve and the centroid to obtain a sequence L1 and a sequence L3 respectively; adding the centroid after the sequence L1 to obtain a first closed curve coordinate sequence; adding the starting point of the sequence L1, the ending point of the sequence L1 and the centroid after the sequence L3 to obtain a second closed curve coordinate sequence;

[0027] Step six: calculating a first closed area according to the first closed curve coordinate sequence by a Gaussian shoelace formula, calculating a second closed area according to the second closed curve coordinate sequence by the Gaussian shoelace formula, and obtaining an edge wear area by subtracting the first closed area from the second closed area;

[0028] Step seven: judging whether the edge wear area is greater than the set threshold, if yes, then performing an alarm.

[0029] Further, the method for converting the obtained image information into a gray image is:

[0030] grayImage = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

[0031] wherein grayImage is a gray image matrix, img is picture information, cv2.cvtColor is a conversion function, and cv2.COLOR_BGR2GRAY is a conversion parameter.

[0032] Specifically, the picture information of the last-stage blade of the low-pressure cylinder of the steam turbine is obtained by using a high-speed camera or a high-speed video camera, the picture information is img, the img is converted into a gray image grayImage, and the method used is the cvtColor package in openCV; the gray image grayImage is a two-dimensional matrix with a horizontal length of h and a vertical length of v, and h and v are related to the size of the original image.

[0033] Further, the method for obtaining the image matrix R of the gray image obtained in step one includes:

[0034] Step two one: performing a box filter on the gray image to obtain an image matrix L;

[0035] L = cv2.boxFilter(grayImage, -1, (F_1, F_2), normalize = 0)

[0036] wherein cv2.boxFilter is a box filter function, grayImage is a gray image, normalize = 0 indicates that no normalization processing is performed, (F_1, F_2) is a kernel of the box filter, F_1 = F_2, and the values of F_1 and F_2 are in the range of 2 to 5;

[0037] Step two two: performing gradient processing on the obtained image matrix L to obtain an image matrix P;

[0038] Specifically, the gray image grayImage is subjected to box filtering and gradient processing, so as to eliminate noise points and expose the metal edges of the blade.

[0039] Step two three: performing erosion on the image matrix P to obtain an image matrix Q;

[0040] Q = cv2.erode(P, D, k1)

[0041] wherein cv2.erode is an erosion function, D is a kernel, k1 is the number of corrosion iterations, k1 is in the range of 1-2;

[0042] Step two four: dilate the image matrix Q to obtain the image matrix R;

[0043] R = cv2.dilate(Q, D, k2)

[0044] Wherein, cv2.dilate is the dilate function, k2 is the number of dilate iterations, k2 is in the range of 1-2.

[0045] Specifically, the erosion and dilatation processing is carried out, the purpose is to make the edge closed; the image matrix P is eroded to obtain the image matrix Q, the method is erode function in opencv(cv2), the image matrix Q is eroded to obtain the image matrix R, the method is dilate function in opencv(cv2), the eroded and dilated image is as shown in Figure 4 .

[0046] Further, the obtained image matrix L is gradient processed to obtain the image matrix P, including:

[0047] According to the graph matrix L, the edge matrix sobelxy is obtained by sobel operator, the obtained edge matrix sobelxy is binarized to obtain the image matrix P;

[0048] sobelxy = cv2.Sobel(src = L, ddepth = cv2.CV_64F, dx = dx1, dy = dy1, ksize = z)

[0049] Wherein, cv2.Sobel is the edge detection function, src is the calculation object, ddepth = cv2.CV_64F is the calculation depth, dx = dx1, dy = dy1 is the first derivative, ksize is the size of sobel operator, z is 9 or 11;

[0050]

[0051] Wherein, a = [1, 2... h], b = [1, 2... v], h is the horizontal length of the gray image, v is the vertical length of the gray image, a is the row sequence of the image matrix P, and b is the column sequence of the image matrix P.

[0052] Specifically, the gradient of the graph matrix L is calculated using the sobel method using the Sobel function of openCV. The calculated matrix is sobelxy; the calculated gradient has negative numbers, so the calculated gradient is binarized, and the edges are assigned black. The processing method is to process the matrix sobelxy, when the element is 0, it becomes 255, if it is not 0, it becomes 0. The processed matrix is P.

[0053] Further, the edge curve of the leaf is obtained according to the image matrix R, and the convex hull line and the centroid of the leaf are obtained according to the edge curve of the leaf, including:

[0054] Step two five: the edge curve H of the leaf is obtained according to the image matrix R through the findContours function;

[0055] H = findContours(R, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)

[0056] Wherein, cv2.RETR_EXTERNAL and cv2.CHAIN_APPROX_NONE) are parameters;

[0057] Step two six: the convex hull line J is obtained according to the edge curve H through the cv2.convexHull function;

[0058] J = cv2.convexHull(H)

[0059] Step two seven: the centroid A[l,m] of the leaf is obtained according to the edge curve H;

[0060]

[0061] Wherein, max(·) is the maximum value, min(·) is the minimum value, is rounded up, c is the row vector of the edge curve, d is the column vector of the edge curve, l is the row coordinate of the centroid, and m is the column coordinate of the centroid.

[0062] Further, the calculation method of the angle between the convex hull line and the centroid is:

[0063] AnglePoint_T = math.atan2(-(e-l), (f-m))*180 / 3.14 wherein, AnglePoint_T is the angle between the convex hull line and the centroid; math.atan is an angle calculation function, e is the row vector of the convex hull line, and f is the column vector of the convex hull line;

[0064] AnglePoint_T represents the change curve of the angle when the centroid moves counterclockwise along the convex hull line.

[0065] The calculation method of the included angle of the edge curve and the centroid is:

[0066] AnglePoint_r = math.atan2(-(c-l), (d-m)) * 180 / 3.14, wherein AnglePoint_r is the included angle of the edge curve and the centroid, c is the row vector of the edge curve, and d is the column vector of the edge curve;

[0067] AnglePoint_r is the change curve of the angle when the centroid moves along the Hundred Birds Garden curve counterclockwise.

[0068] The calculation method of the distance of the edge curve and the centroid is:

[0069]

[0070] wherein LengthPoint_r is the distance of the edge curve and the centroid;

[0071] Specifically, LengthPoint_r represents the distance change curve when the centroid moves along the edge curve counterclockwise.

[0072] Further, the angle and distance data are filtered and extreme value screened to obtain the sequence number of the included angle sequence of the blade edge starting point and the included angle sequence of the blade edge ending point of the edge curve and the centroid, including:

[0073] Step four one: selecting the point with an angle greater than Angle1 and less than Angle2 in the included angle sequence AnglePoint_r of the edge curve and the centroid, and obtaining the index starting sequence number H_low and the index ending sequence number H_up according to the selected point;

[0074] Step four two: truncating the distance sequence LengthPoint_r of the edge curve and the centroid according to the index starting sequence number H_low and the index ending sequence number H_up to obtain a new sequence LPJ, filtering LPJ according to the first window size to obtain an LPF signal, finding the peak point LPP of the LPF signal, and obtaining the coordinate index [LPP_low, LPP_up] according to the LPP;

[0075] LPF = savgol_filter(LPJ, window1, Deg)

[0076] wherein savgol_filter is a filtering function 1, window1 is the first window size, window1 is selected from the range of 600 to 800, Deg is the polynomial order, and Deg = 3;

[0077] LPP=argrelextrema(LPF,np.greater)

[0078] Where argrelextrema is the function for finding extrema, and np.greater means finding the maximum value;

[0079] Step 43: Filter LPJ according to the second window size to obtain the LPF1 signal, and find the peak point of the LPF1 signal to obtain LPP1; find the points in LPP1 with coordinate indices [LPP_low, LPP_up] to form the sequence LPP. c According to the LPP sequence c Get LPP c Minimum value of LPP clow and LPP c maximum value of LPP cup ;

[0080] LPF1=savgol_filter(LPJ,window2,Deg)

[0081] LPP1=argrelextrema(LPF1,np.greater)

[0082] window2 is the size of the second window, and its value is between 40 and 50.

[0083] Step 44: According to LPP c Minimum value of LPP clow LPP c maximum value of LPP up The sequence number Node is obtained by using the coordinate index [LPP_low, LPP_up] to determine the angle sequence between the blade edge starting point and the edge curve and the centroid. low The sequence number of the Node, which is the angle between the blade edge endpoint and the edge curve and the centroid. up ;

[0084] Node low =LPP clow +LPP_low

[0085] Node up =LPP cup +LPP_low.

[0086] Specifically, LPP clow It is the sequence number of the starting point of the blade edge, LPP cupis the sequence number of the leaf edge end point, the sequence number of the corresponding point is consistent in LengthPoint_r and AnglePoint_r, the sequence number of the leaf edge start point in LengthPoint_r and AnglePoint_r is Node low , and the sequence number of the leaf edge end point in LengthPoint_r and AnglePoint_r is Node up .

[0087] Further, the value range of Angle1 is [-135, -90], and the value range of Angle2 is [45, 60].

[0088] Further, according to the sequence number of the leaf edge start point in the sequence of the angle between the edge curve and the centroid and the sequence number of the leaf edge end point in the sequence of the angle between the edge curve and the centroid, the sequence L1 and the sequence L3 are obtained by intercepting the sequence of the angle between the edge curve and the centroid and the sequence of the distance between the convex hull line and the centroid, and the sequence L1 and the sequence L3 include:

[0089] Step five: find the sequence number Node low and the sequence number Node up in the sequence of the angle between the edge curve and the centroid AnglePoint_r corresponding to the angle AnglePoint_r[Node low ] and AnglePoint_r[Node up ];

[0090] Step five two: truncate the sequence of the distance between the edge curve and the centroid LengthPoint_r according to the sequence number [Node low , Node up ] to obtain the sequence L1, the start point of the sequence L1 is LengthPoint_r[Node low ], and the end point is LengthPoint_r[Node up ];

[0091] Find the sequence L3 in the sequence of the angle between the convex hull line and the centroid AnglePoint_T whose angle is between AnglePoint_r[Node low ] and AnglePoint_r[Node up ];

[0092] Add the centroid after the sequence L1 to obtain the first closed curve coordinate sequence, and add the start point of the sequence L1, the end point of the sequence L1 and the centroid after the sequence L3 to obtain the second closed curve coordinate sequence, which includes:

[0093] Step five three: add the centroid [l, m] after L1 to form a new sequence L2, L2(xi y i ), i∈[1, s1], s1 is the data length of the sequence L2, L2 is the first closed curve coordinate sequence, x i is the row coordinate of the sequence L2, y i is the column coordinate of the sequence L2.

[0094] The sequence L3, LengthPoint_r[Node low ], LengthPoint_r[Node up ] and the centroid A[l, m] are arranged in sequence to obtain a new sequence L4, L4(g j h j ), j∈[1, s2], s2 is the data length of the sequence L4, L4 is the second closed curve coordinate sequence, g j is the row coordinate of the sequence L4, h j is the column coordinate of the sequence L4.

[0095] Further, the first closed area is calculated according to the first closed curve coordinate sequence by the Gauss shoelace formula, the second closed area is calculated according to the second closed curve coordinate sequence by the Gauss shoelace formula, and the edge wear area is obtained by subtracting the first closed area from the second closed area, comprising:

[0096] The first closed area A1 is calculated according to the first closed curve coordinate sequence L2 by the Gauss shoelace formula;

[0097]

[0098] The second closed area A2 is calculated according to the second closed curve coordinate sequence L4 by the Gauss shoelace formula;

[0099]

[0100] The edge wear area A cal is obtained by subtracting the first closed area A1 from the second closed area A2, A cal =A2-A1.

[0101] When the area A cal exceeds a set threshold value, an alarm is triggered, indicating that the current blade edge wear is too large and needs to be processed, and the picture information and intermediate calculation process of the last stage blade of the steam turbine are stored.

[0102] While the application has been described with reference to particular embodiments thereof, it is to be understood that these embodiments are merely illustrative of the principles and applications of the present application. It will be apparent to those skilled in the art that numerous modifications can be made within the scope of the present application as defined by the appended claims. It is intended that all such modification fall within the spirit and scope of the present application. It will be understood that the features described in connection with one embodiment can be used in connection with another embodiment.

Claims

1. A method for identifying edge wear of a last stage blade of a low pressure cylinder of a steam turbine, characterized by, The method comprises the following steps: Step one: obtaining image information of a last-stage blade of a steam turbine, and converting the obtained image information into a gray image; Step two: obtaining an image matrix R of the gray image obtained in step one; obtaining an edge curve according to the image matrix R, and obtaining a convex hull line and a centroid according to the edge curve; Step three: calculating an included angle and distance data according to the edge curve, the convex hull line and the centroid, wherein the included angle and distance data comprise an included angle sequence of the convex hull line and the centroid, an included angle sequence of the edge curve and the centroid, and a distance sequence of the edge curve and the centroid; Step four: filtering and extreme value screening the included angle and distance data to obtain a sequence number of a blade edge starting point in the included angle sequence of the edge curve and the centroid and a sequence number of a blade edge ending point in the included angle sequence of the edge curve and the centroid; Step five: cutting the included angle sequence of the edge curve and the centroid and the distance sequence of the convex hull line and the centroid according to the obtained sequence number of the blade edge starting point in the included angle sequence of the edge curve and the centroid and the sequence number of the blade edge ending point in the included angle sequence of the edge curve and the centroid to obtain a sequence L1 and a sequence L3 respectively; adding the centroid after the sequence L1 to obtain a first closed curve coordinate sequence; adding a starting point of the sequence L1, an ending point of the sequence L1 and the centroid after the sequence L3 to obtain a second closed curve coordinate sequence; Step six: calculating a first closed area according to the first closed curve coordinate sequence by using a Gaussian shoelace formula, calculating a second closed area according to the second closed curve coordinate sequence by using the Gaussian shoelace formula, and obtaining an edge wear area by subtracting the first closed area from the second closed area; Step seven: judging whether the edge wear area is greater than a set threshold value, and if yes, alarming; The method for obtaining the edge curve of the blade according to the image matrix R, and obtaining the convex hull line and the centroid of the blade according to the edge curve of the blade comprises the following steps: Step two five: obtaining the edge curve H of the blade according to the image matrix R by using a findContours function; H = findContours(R, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE) Wherein, cv2.RETR_EXTERNAL and cv2.CHAIN_APPROX_NONE are parameters of the findContours function; Step two six: obtaining the convex hull line J according to the edge curve H of the blade by using a cv2.convexHull function; J = cv2.convexHull(H) Step two seven: obtaining the centroid A[l,m] according to the edge curve H of the blade; where max( ) is the maximum value and min( ) is the minimum value, is the ceiling function, c is the edge curve row vector, d is the edge curve column vector, l is the row coordinate of the centroid, and m is the column coordinate of the centroid.

2. A method for identifying edge wear of a last stage blade of a low pressure cylinder of a steam turbine according to claim 1, characterized in that: The method for converting the obtained image information into a gray image comprises the following steps: grayImage = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY) Wherein, grayImage is a gray image matrix, img is picture information, cv2.cvtColor is a conversion function, and cv2.COLOR_BGR2GRAY is a conversion parameter.

3. The method for identifying the edge wear of the last stage blade of the low-pressure cylinder of a steam turbine according to claim 1, characterized in that: The method for obtaining the image matrix R of the gray image obtained in step one comprises the following steps: Step two one: performing square filtering on the gray image to obtain an image matrix L; L = cv2.boxFilter(grayImage, -1, (F_1, F_2), normalize = 0) Wherein, cv2.boxFilter is a box filter function, grayImage is a gray image, normalize = 0 indicates that no normalization processing is performed, (F_1, F_2) is the kernel of the box filter, F_1 = F_2, and the values of F_1 and F_2 are in the range of 2 to 5; Step two: the obtained image matrix L is subjected to gradient processing to obtain an image matrix P; Step two three: the image matrix P is subjected to corrosion to obtain an image matrix Q; Q = cv2.erode (P, D, k1) wherein cv2.erode is an erosion function, k1 is the number of erosion iterations, k1 is in the range of 1-2; Step two four: the image matrix Q is subjected to inflation to obtain an image matrix R; R = cv2.dilate (Q, D, k2) Wherein, cv2.dilate is an inflation function, k2 is the inflation iteration number, and the value of k2 is in the range of 1-2.

4. A method of identifying edge wear of a last stage blade of a low pressure cylinder of a steam turbine according to claim 3, characterized in that: The obtained image matrix L is subjected to gradient processing to obtain an image matrix P, including: According to the graph matrix L, an edge matrix sobelxy is obtained by a sobel operator, and the obtained edge matrix sobelxy is binarized to obtain an image matrix P; sobelxy = cv2.Sobel (src = L, ddepth = cv2.CV_64F, dx = dx1, dy = dy1, ksize = z) Wherein, cv2.Sobel is an edge detection function, src = L is the calculation object, ddepth = cv2.CV_64F is the calculation depth, dx = dx1, dy = dy1 is the first derivative, ksize is the size of the sobel operator, and z is 9 or 11; Wherein, a = [1, 2...h], b = [1, 2...v], h is the horizontal length of the gray image, v is the vertical length of the gray image, a is the row sequence of the image matrix P, and b is the column sequence of the image matrix P.

5. The method according to claim 1, wherein: The calculation method of the angle sequence between the convex hull line and the centroid is: AnglePoint_T = math.atan2(-(e-l), (f-m)) * 180 / 3.14 Wherein, AnglePoint_T is the angle sequence between the convex hull line and the centroid; math.atan is an angle calculation function, e is a row vector of the convex hull line, and f is a column vector of the convex hull line; The calculation method of the angle sequence between the edge curve and the centroid is: AnglePoint_r = math.atan2(-(c-l), (d-m)) * 180 / 3.14 Wherein, AnglePoint_r is the angle sequence between the edge curve and the centroid, c is a row vector of the edge curve, and d is a column vector of the edge curve; The calculation method of the distance sequence between the edge curve and the centroid is: Wherein, LengthPoint_r is the distance sequence between the edge curve and the centroid.

6. A method of identifying edge wear of a last stage blade of a low pressure cylinder of a steam turbine according to claim 5, characterized in that: Filtering and extreme value screening of the included angle and distance data to obtain the sequence number of the included angle sequence of the leaf edge starting point between the edge curve and the centroid and the sequence number of the included angle sequence of the leaf edge ending point between the edge curve and the centroid, comprising: Step four one: selecting the points in the included angle sequence AnglePoint_r between the edge curve and the centroid whose angles are greater than Angle1 and less than Angle2, and obtaining the index starting sequence number H_low and the index ending sequence number H_up according to the selected points; Step four two: truncating the distance sequence LengthPoint_r between the edge curve and the centroid according to the index starting sequence number H_low and the index ending sequence number H_up to obtain a new sequence LPJ, filtering LPJ according to the first window size to obtain an LPF signal, finding the peak point LPP of the LPF signal, and obtaining the coordinate index [LPP_low, LPP_up] according to the LPP; LPF = savgol_filter(LPJ, window1, Deg) wherein savgol_filter is a filter function 1, window1 is a first window size, window1 is selected from the range of 600 to 800, Deg is a polynomial order, and Deg = 3; LPP = argrelextrema(LPF, np.greater) wherein argrelextrema is an extreme value finding function, and np.greater indicates finding a maximum value; Step 43: Filter LPJ according to the second window size to obtain the LPF1 signal, and find the peak point of the LPF1 signal to obtain LPP1; find the points in LPP1 with coordinate indices [LPP_low, LPP_up] to form the sequence LPP. c According to the LPP sequence c Get LPP c Minimum value of LPP clow and LPP c maximum value of LPP cup ; LPF1 = savgol_filter(LPJ, window2, Deg) LPP1 = argrelextrema(LPF1, np.greater) window2 is a second window size, window2 is selected from the range of 40 to 50, Step four four: The leaf edge start point Node c is obtained according to the minimum value LPP clow of LPP c , the maximum value LPP cup of LPP low and the coordinate index [LPP_low, LPP_up] up ; Node low = LPP clow + LPP_low Node up = LPP cup + LPP_low.

7. A method of identifying edge wear of a last stage blade of a low pressure cylinder of a steam turbine according to claim 6, characterized in that: According to the obtained sequence number of the included angle sequence of the leaf edge starting point between the edge curve and the centroid and the sequence number of the included angle sequence of the leaf edge ending point between the edge curve and the centroid, the included angle sequence between the edge curve and the centroid and the distance sequence between the convex hull and the centroid are truncated to obtain sequences L1 and L3, comprising: Step five: find the sequence of the angle between the edge curve and the centroid AnglePoint_r[Node low ] and AnglePoint_r[Node up ]; the corresponding angle AnglePoint_r[Node low ] and AnglePoint_r[Node up ]; Step five two: the sequence LengthPoint_r of the distance between the edge curve and the centroid is truncated according to the sequence number [Node low , Node up ] to obtain sequence L1, the starting point of sequence L1 is LengthPoint_r[Node low ], and the ending point is LengthPoint_r[Node up ]. Find the sequence of angles between AnglePoint_r[Node low ] and AnglePoint_r[Node up ] in AnglePoint_T and store it in L3; Adding the centroid after sequence L1 to obtain a first closed curve coordinate sequence; adding the starting point of sequence L1, the ending point of sequence L1, and the centroid after sequence L3 to obtain a second closed curve coordinate sequence, comprising: Step five three: add the centroid [l,m] after L1 to form a new sequence L2, L2(x i ,y i ), i∈[1,s1], s1 is the data length of sequence L2, L2 is the first closed curve coordinate sequence, x i is the row coordinate of sequence L2, and y i is the column coordinate of sequence L2; L3, LengthPoint_r[Node low ], LengthPoint_r[Node up ] and the centroid A[l,m] are arranged in sequence to obtain a new sequence L4, L4(g j ,h j ), j∈[1,s2], s2 is the data length of the sequence L4, L4 is the second closed curve coordinate sequence, g j is the row coordinate of the sequence L4, and h j is the column coordinate of the sequence L4.

8. A method of identifying edge wear of a last stage blade of a low pressure cylinder of a steam turbine according to claim 7, characterized in that: According to the first closed curve coordinate sequence, a first closed area is calculated by the Gauss shoelace formula, according to the second closed curve coordinate sequence, a second closed area is calculated by the Gauss shoelace formula, and the edge wear area is obtained by subtracting the first closed area from the second closed area, comprising: According to the first closed curve coordinate sequence L2, a first closed area A1 is calculated by the Gauss shoelace formula; According to the second closed curve coordinate sequence L4, a second closed area A2 is calculated by the Gauss shoelace formula; The edge wear area A is obtained by differencing the first enclosed area Al and the second enclosed area A2 cal , A cal = A2 - Al.

9. A method of identifying edge wear of a last stage blade of a low pressure cylinder of a steam turbine according to claim 8, characterized in that: The value range of Angle1 is [-135, -90], and the value range of Angle2 is [45, 60].

Images