A method for preparing a high-precision high-temperature alloy spring

Abstract

This invention relates to the field of spring manufacturing technology, specifically disclosing a method for manufacturing a high-precision high-temperature alloy spring. The method includes: selecting a specific material, processing the spring through three heat treatments combined with appropriate heat preservation and cooling methods; drawing the heat-treated alloy blank to a wire diameter of 0.1-3 mm; initial drawing at room temperature; precision drawing at low temperature; and then winding the spring using a CNC spring forming machine, controlling the tensile force and winding speed to ensure dimensional standards; surface polishing and chrome plating for protection; calculating the spring length difference and related ratios to obtain the spring precision Jd by detecting the spring length; generating an anomaly analysis signal by comparing the spring precision Jd with a threshold Jdy; plotting a temperature-spring precision Jd curve based on the anomaly analysis signal; calculating the anomaly degree coefficient Cx; and finally establishing a model using a clustering algorithm to calculate the optimal temperature. Through monitoring and optimization of the spring, the spring precision and quality are improved.

Description

Technical Field

[0001] This invention relates to the field of spring manufacturing technology, and specifically to a method for manufacturing a high-precision high-temperature alloy spring. Background Technology

[0002] In modern industry, springs are an important elastic element and are widely used in many fields. With the development of industry, the performance requirements for springs are becoming increasingly stringent, especially for springs in high-temperature working environments. They not only need to have good high-temperature resistance, but also need to maintain high precision to ensure the stable operation of equipment.

[0003] A Chinese patent application with publication number CN105921649A discloses a method for preparing a high-precision high-temperature alloy spring, comprising: (1) firstly, making a variable-diameter mandrel with a working coil diameter smaller than the diameter of its two side support coils, and setting a wire-stopping pin on the outside of one side support coil; (2) clamping the variable-diameter mandrel on a lathe, clamping the high-temperature alloy steel wire with a wire clamping plate, and feeding the wire end into the wire-stopping pin; (3) cutting off the excess steel wire to ensure the total number of coils and the effective number of coils, and then loading it into a heat treatment furnace for strengthening treatment; (4) grinding the two ends of the spring on a grinding wheel to ensure that the spring end coil is ground to 3 / 4. After each grinding, cooling in water is performed once. After grinding, the spring height is checked to see if it reaches the specified height; (5) making a pressure sleeve that matches the pressure shaft, and setting a disc base integral with it at the bottom of the pressure shaft; (6) clamping the pressure sleeve on a lathe, and mounting the ground spring on the pressure shaft. Moving the tailstock of the lathe so that it is against the center of the disc base. After the spring is compressed, stop moving the tailstock and hold for at least 10 seconds before releasing. Repeat this process until the spring force value meets the usage requirements.

[0004] In the existing technology, the traditional spring manufacturing method has many limitations in the production of high-precision high-temperature alloys. On the one hand, the material selection is not optimized enough and it is difficult to meet the requirements of high temperature resistance and high precision. On the other hand, there is a lack of effective means to control the spring precision. In the heat treatment process, the precision of the spring is detected and evaluated, and the optimal temperature range is selected to control the spring precision, so as to improve the reliability and durability of the spring in practical applications.

[0005] Therefore, the present invention provides a method for preparing a high-precision high-temperature alloy spring. Summary of the Invention

[0006] The purpose of this invention is to provide a method for preparing a high-precision high-temperature alloy spring, so as to solve the problems mentioned above.

[0007] The objective of this invention can be achieved through the following technical solutions:

[0008] Step 1: Obtain the spring length data, calculate the spring length difference with the target spring length value, mark springs whose spring length difference does not meet the threshold as abnormal springs, calculate the average length difference of abnormal springs and calculate with the spring length difference threshold to obtain the abnormal length ratio Cd, calculate the proportion of abnormal springs to obtain the abnormal spring number ratio SI, and calculate the spring accuracy Jd based on the abnormal spring number ratio SI and abnormal length ratio Cd. If the spring accuracy Jd ≥ the spring accuracy threshold Jdy, generate the abnormal analysis signal.

[0009] Step 2: Based on the anomaly analysis signal, collect data from multiple production batches to obtain the change in spring accuracy Jd under different temperature conditions, plot the temperature-spring accuracy Jd curve, plot a reference line using the spring accuracy threshold Jdy as a reference value, calculate the abnormal curve segment where the spring accuracy Jd exceeds the spring accuracy threshold Jdy, obtain the abnormal area enclosed by the abnormal curve segment and the reference line, obtain the temperature block length value corresponding to the abnormal curve segment, calculate the average abnormal area ratio Yc and the temperature block length ratio Wd, and calculate the anomaly degree coefficient Cx. The anomaly degree coefficient Cx ≥ the anomaly degree coefficient threshold Cxz, and generate the degree analysis signal.

[0010] Step 3: Based on the degree analysis signal, obtain the spring accuracy Jd corresponding to different temperatures, use a clustering algorithm to establish a clustering algorithm model of temperature-spring accuracy Jd, and calculate the optimal temperature.

[0011] As a further technical solution of the present invention: the method for obtaining the spring accuracy Jd is as follows:

[0012] The spring accuracy Jd is calculated based on the abnormal spring quantity ratio SI and the abnormal length ratio Cd.

[0013] Through the formula: Obtain the spring accuracy Jd, where a and b are preset proportional coefficients;

[0014] The abnormal analysis signal is generated in the following way:

[0015] Compare the spring accuracy Jd with the spring accuracy threshold Jdy;

[0016] If the spring accuracy Jd is greater than or equal to the spring accuracy threshold Jdy, an anomaly analysis signal is generated.

[0017] As a further technical solution of the present invention: the method for obtaining the number ratio SI of abnormal springs is as follows: obtain the length data of the spring sample, and subtract the spring sample from the target value of the spring length to obtain the spring length difference;

[0018] Compare the spring length difference with the spring length difference threshold;

[0019] It should be noted that the spring length difference threshold is set by professionals in the field based on experience, and the target value of the spring length is the expected value for spring production.

[0020] Springs whose spring length difference does not fall within the threshold range are marked as abnormal springs;

[0021] Obtain the number of abnormal springs, and calculate the ratio of the number of abnormal springs to the total number of springs tested to obtain the abnormal spring count ratio. Mark the abnormal spring count ratio as Sl.

[0022] As a further technical solution of the present invention: the abnormal length ratio Cd is obtained as follows:

[0023] The critical spring length difference is obtained by subtracting the length difference of the abnormal spring from the endpoint of the nearest adjacent spring length difference threshold range and taking the absolute value.

[0024] Obtain the critical length difference of all abnormal springs, sum the critical length differences of all abnormal springs, and get the total critical length difference value.

[0025] The average critical difference of spring length is obtained by comparing the total difference in the critical length of the springs with the total number of all abnormal springs.

[0026] The abnormal length ratio is obtained by comparing the average critical difference of spring length with the length of the spring length difference threshold range. The abnormal length ratio is then labeled as Cd.

[0027] As a further technical solution of the present invention: the abnormality coefficient Cx is obtained as follows:

[0028] The anomaly degree coefficient Cx is calculated based on the average anomaly area ratio Yc and the temperature block length ratio Wd.

[0029] Through the formula: The abnormality coefficient Cx was calculated.

[0030] As a further technical solution of the present invention: the method for obtaining the average abnormal area ratio Yc is as follows:

[0031] Plot the temperature-spring accuracy Jd curve in a two-dimensional rectangular coordinate system with temperature as the X-axis and spring accuracy Jd as the Y-axis.

[0032] Using the spring accuracy threshold Jdy as a reference value, draw a horizontal reference line in the coordinate system;

[0033] Obtain the curve segment where the spring accuracy Jd exceeds the spring accuracy threshold Jdy; this is the abnormal curve segment.

[0034] The area enclosed by a single abnormal curve segment and the reference line is used to obtain the abnormal area.

[0035] Obtain the area enclosed by a single abnormal curve segment and the X-axis to get the curve area;

[0036] The ratio of the abnormal area to the curve area is calculated to obtain the abnormal area ratio.

[0037] Obtain the ratio of abnormal area for all abnormal curve segments and sum them to get the total ratio of abnormal area;

[0038] The ratio of the total abnormal area to the number of abnormal curve segments is used to obtain the average abnormal area ratio, which is then labeled as Yc.

[0039] As a further technical solution of the present invention: the method for obtaining the temperature block length ratio Cd is as follows:

[0040] Obtain the abnormal temperature segment corresponding to the abnormal curve segment on the X-axis, and calculate the length of the abnormal temperature segment;

[0041] The ratio of the abnormal temperature segment length to the total temperature segment length is calculated to obtain the temperature block length ratio, which is then denoted as Wd.

[0042] As a further technical solution of the present invention: the method for generating the degree analysis signal is as follows:

[0043] Compare the anomaly severity coefficient Cx with the anomaly severity coefficient threshold Cxz;

[0044] If the anomaly degree coefficient Cx ≥ the anomaly degree coefficient threshold Cxz, a degree analysis signal is generated.

[0045] As a further technical solution of the present invention: the optimal temperature is obtained by means of:

[0046] Based on the trained temperature-spring accuracy Jd clustering algorithm model, the mean spring accuracy within each cluster is calculated over the temperature range.

[0047] Obtain the average precision of all clusters, sort the average precision in descending order, and obtain the highest precision;

[0048] Find the temperature range of the cluster with the highest accuracy, and then obtain the optimal temperature range;

[0049] The optimal temperature is obtained by summing and averaging the optimal temperature ranges.

[0050] As a further technical solution of the present invention: the construction method of the temperature-spring accuracy Jd clustering algorithm model is as follows:

[0051] SS1. Perform data preprocessing on the acquired temperature-spring accuracy Jd set H, removing outliers and standardizing the data;

[0052] SS2. Select the K-Means clustering algorithm, use the elbow rule to determine the number K, and set K=8 as the number of clusters;

[0053] SS3: Randomly initialize three cluster centers, calculate the distance between the data point and the three cluster centers, treat the data pair as the data point within the cluster center, and assign the data point to the cluster containing the nearest cluster center;

[0054] SS4. Recalculate the center of each cluster, i.e. the mean of all data points in the cluster. Repeat step SS3 until the cluster centers no longer change or the predetermined 200 iterations are reached.

[0055] The beneficial effects of this invention are:

[0056] (1) A high-temperature resistant nickel-based alloy is selected, and aluminum, molybdenum and niobium elements are added to improve the high-temperature resistance of the spring. In the heat treatment process, the high-temperature resistance and metal fatigue resistance of the spring are increased through three heat treatment processes, so that it can maintain stable performance in high-temperature environment. The surface of the spring is polished to make the spring surface smooth and flat, which further improves the performance stability of the spring. The surface of the spring is chrome-plated to improve the oxidation and corrosion of the spring in high-temperature environment and corrosive medium, which effectively improves the reliable operation capability of the spring under load conditions.

[0057] (2) By acquiring the spring length data and calculating the spring length difference with the target spring length value, the ratio of abnormal springs and the ratio of abnormal lengths are calculated. The spring accuracy (jd) is also calculated. By analyzing the variation of spring accuracy (Jd) under different temperatures, a temperature-spring accuracy (Jd) curve is plotted. Parameters related to the abnormal curve segments are calculated to assess the degree of abnormality of the springs at different temperatures. A temperature-spring accuracy (Jd) clustering algorithm model is established using a clustering algorithm. The model can calculate the average spring accuracy within each cluster, thereby determining the optimal temperature range and optimal temperature. By calculating the optimal temperature, the springs can function stably in applications with extremely high accuracy requirements, thus improving the overall operating performance and stability of the equipment system. Attached Figure Description

[0058] The invention will now be further described with reference to the accompanying drawings.

[0059] Figure 1 This is a flowchart of a method for preparing a high-precision high-temperature alloy spring according to the present invention;

[0060] Figure 2 This is a flowchart of a spring detection method according to the present invention. Detailed Implementation

[0061] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] Example 1

[0063] Please see Figure 1 As shown, this invention relates to a method for preparing a high-precision high-temperature alloy spring.

[0064] S1. High-temperature resistant nickel-based alloys are selected as raw materials, and aluminum, molybdenum, and niobium elements are added to the materials;

[0065] S2. First heat treatment: Heat the alloy to 1100℃-1150℃ and hold for 1-2 hours, then cool the alloy rapidly. Second heat treatment: Heat to 650℃-700℃, hold for 4-8 hours, then cool slowly. Third heat treatment: Heat to 700℃-750℃, hold for 6-10 hours, then cool.

[0066] S3. The heat-treated alloy billet is processed to the required wire diameter through a multi-pass drawing machine. The wire diameter range is between 0.1 and 3 mm. The initial drawing is carried out at room temperature, and the precision drawing is carried out at low temperature.

[0067] S4. Use a CNC spring forming machine to wind the wire into shape;

[0068] S5. Polish the spring surface to remove residual oxides from the processing, and plate chromium onto the spring surface as a high-temperature resistant protective coating.

[0069] Example 2

[0070] Based on Example 1, a method for preparing a high-precision high-temperature alloy spring further includes:

[0071] Step 1: After completing step S5, obtain the spring length data, calculate the spring length difference with the target spring length value, mark springs whose spring length difference does not meet the threshold as abnormal springs, calculate the average length difference of abnormal springs with the spring length difference threshold to obtain the abnormal length ratio Cd, calculate the proportion of abnormal springs to obtain the abnormal spring quantity ratio SI, and calculate the spring accuracy Jd based on the abnormal spring quantity ratio SI and the abnormal length ratio Cd. If the spring accuracy Jd ≥ the spring accuracy threshold Jdy, generate an abnormality analysis signal.

[0072] Select the spring prepared in step S5 as the test sample and mark the test sample as the spring sample;

[0073] Obtain the length data of the spring sample, and subtract the spring length of the spring sample from the target spring length value to obtain the spring length difference;

[0074] Compare the spring length difference with the spring length difference threshold range;

[0075] It should be noted that the spring length difference threshold range is set by professionals in the field based on experience, and the target value of the spring length is the expected value for spring production.

[0076] Springs whose spring length difference does not fall within the threshold range are marked as abnormal springs;

[0077] Obtain the number of abnormal springs, and calculate the ratio of the number of abnormal springs to the total number of springs tested to obtain the abnormal spring count ratio. Mark the abnormal spring count ratio as S l.

[0078] The critical spring length difference is obtained by subtracting the length difference of the abnormal spring from the endpoint of the nearest adjacent spring length difference threshold range and taking the absolute value.

[0079] Obtain the critical length difference of all abnormal springs, sum the critical length differences of all abnormal springs, and get the total critical length difference value.

[0080] The average critical difference of spring length is obtained by comparing the total difference in the critical length of the springs with the total number of all abnormal springs.

[0081] The ratio of the average critical difference of spring length to the length of the threshold range of spring length difference is processed to obtain the abnormal length ratio, which is marked as Cd.

[0082] The spring length difference threshold range length value is obtained by taking the absolute value of the difference between the two endpoints of the spring length difference threshold range.

[0083] The spring accuracy Jd is calculated based on the abnormal spring quantity ratio SI and the abnormal length ratio Cd.

[0084] Through the formula: Obtain the spring accuracy Jd, where a and b are preset proportional coefficients;

[0085] Compare the spring accuracy Jd with the spring accuracy threshold Jdy;

[0086] If the spring accuracy Jd ≥ the spring accuracy threshold Jdy, it indicates that the spring accuracy varies greatly, generating an anomaly analysis signal;

[0087] If the spring accuracy Jd < the spring accuracy threshold Jdy, it indicates that the spring accuracy difference is within the expected range, and it is still necessary to continue monitoring the abnormal coefficient changes of the spring.

[0088] Step 2: Based on the anomaly analysis signal, collect data from multiple production batches to obtain the change in spring accuracy Jd under different temperature conditions, plot the temperature-spring accuracy Jd curve, plot a reference line using the spring accuracy threshold Jdy as a reference value, calculate the abnormal curve segment where the spring accuracy Jd exceeds the spring accuracy threshold Jdy, obtain the abnormal area enclosed by the abnormal curve segment and the reference line, obtain the temperature block length value corresponding to the abnormal curve segment, calculate the average abnormal area ratio Yc and the temperature block length ratio Wd, and calculate the anomaly degree coefficient Cx. The anomaly degree coefficient Cx ≥ the anomaly degree coefficient threshold Cxz, and generate the degree analysis signal.

[0089] In the third heat treatment stage, data from multiple production batches were collected to analyze the changes in spring accuracy Jd under different temperature conditions.

[0090] Plot the temperature-spring accuracy Jd curve in a two-dimensional rectangular coordinate system with temperature as the X-axis and spring accuracy Jd as the Y-axis.

[0091] Using the spring accuracy threshold Jdy as a reference value, draw a horizontal reference line in the coordinate system;

[0092] Obtain the curve segment where the spring accuracy Jd exceeds the spring accuracy threshold Jdy; this is the abnormal curve segment.

[0093] The area enclosed by a single abnormal curve segment and the reference line is used to obtain the abnormal area.

[0094] Obtain the area enclosed by a single abnormal curve segment and the X-axis to get the curve area;

[0095] The ratio of the abnormal area to the curve area is calculated to obtain the abnormal area ratio.

[0096] Obtain the ratio of abnormal area for all abnormal curve segments and sum them to get the total ratio of abnormal area;

[0097] The ratio of the total abnormal area to the number of abnormal curve segments is processed to obtain the average abnormal area ratio, and the average abnormal area ratio is marked as Yc.

[0098] Obtain the abnormal temperature segment corresponding to the abnormal curve segment on the X-axis, and calculate the length of the abnormal temperature segment;

[0099] It should be noted that the abnormal temperature range refers to the interval formed between the temperature points corresponding to the two endpoints of the abnormal curve segment on the X-axis.

[0100] The ratio of the abnormal temperature segment length to the total temperature segment length is calculated to obtain the temperature block length ratio, which is then denoted as Wd.

[0101] It should be noted that the total length of the temperature range is the interval formed between the two endpoints of the temperature-spring accuracy Jd curve on the X-axis.

[0102] The anomaly degree coefficient Cx is calculated based on the average anomaly area ratio Yc and the temperature block length ratio Wd.

[0103] Through the formula: The anomaly degree coefficient Cx is calculated, where d and f are preset proportionality coefficients, and ln(f*YC+d*Wd) is the logarithmic function of the base e. f+d It is an exponential function;

[0104] Compare the anomaly severity coefficient Cx with the anomaly severity coefficient threshold Cxz;

[0105] If the abnormality coefficient Cx ≥ the abnormality coefficient threshold Cxz, it indicates that the abnormality coefficient Cx of the spring exceeds the expected value, and a degree analysis signal is generated.

[0106] If the anomaly coefficient Cx < the anomaly coefficient threshold Cxz, it indicates that the anomaly coefficient Cx of the spring is within the expected value, but it is still necessary to continue monitoring the change of the anomaly coefficient threshold Cxz. The technical solution described in this embodiment is as follows: obtain the length data of the spring, calculate the spring length difference with the target spring length, mark the springs whose spring length difference does not meet the threshold as abnormal springs, calculate the average length difference of the abnormal springs and process it with the spring length difference threshold to obtain the abnormal length ratio Cd, calculate the proportion of abnormal springs to obtain the abnormal spring quantity ratio SI, and calculate the spring accuracy Jd based on the abnormal spring quantity ratio SI and the abnormal length ratio Cd. When Jd ≥ spring accuracy threshold Jdy, an anomaly analysis signal is generated. Based on the anomaly analysis signal, data from multiple production batches are collected to obtain the change of spring accuracy Jd under different temperature conditions. A temperature-spring accuracy Jd curve is plotted. A reference line is plotted using the spring accuracy threshold Jdy as a reference value. Anomaly curve segments where spring accuracy Jd exceeds the spring accuracy threshold Jdy are calculated. The anomaly area enclosed by the anomaly curve segment and the reference line is obtained. The temperature block length value corresponding to the anomaly curve segment is obtained. The average anomaly area ratio Yc and the temperature block length ratio Wd are calculated. The anomaly degree coefficient Cx is calculated. The anomaly degree coefficient Cx ≥ the anomaly degree coefficient threshold Cxz, and a degree analysis signal is generated.

[0107] Example 3

[0108] Based on Example 2, the method for preparing a high-precision high-temperature alloy spring further includes:

[0109] Step 3: Based on the degree analysis signal, obtain the spring accuracy Jd corresponding to different temperatures, use a clustering algorithm to establish a clustering algorithm model of temperature-spring accuracy Jd, and calculate the optimal temperature;

[0110] Obtain the spring accuracy Jd corresponding to different temperatures, integrate the temperature and the corresponding spring accuracy Jd into a data pair, and integrate all data pairs into a spring accuracy set H;

[0111] For example, the temperature-spring accuracy Jd data pairs are (700℃, 80%), (750℃, 85%), ...; the spring accuracy set H = {(700℃, 80%), (750℃, 85%), ...};

[0112] A clustering algorithm model for spring temperature-spring accuracy Jd is established based on the clustering algorithm.

[0113] The specific temperature-spring accuracy Jd clustering algorithm model is constructed as follows;

[0114] SS1. Perform data preprocessing on the acquired temperature-spring accuracy Jd set H, removing outliers and standardizing the data;

[0115] SS2. Select the K-Means clustering algorithm, use the elbow rule to determine the number K, and set K=8 as the number of clusters;

[0116] It should be noted that the elbow rule is a technique used to determine the number of data clusters. In the K-Means clustering algorithm, the optimal value of K is determined by analyzing the relationship between the clustering error and the number of clusters K.

[0117] SS3: Randomly initialize three cluster centers, calculate the distance between the data point and the three cluster centers, treat the data pair as the data point within the cluster center, and assign the data point to the cluster containing the nearest cluster center;

[0118] SS4. Recalculate the center of each cluster, i.e. the mean of all data points in the cluster. Repeat step SS3 until the cluster centers no longer change or the predetermined 200 iterations are reached.

[0119] Based on the trained temperature-spring accuracy Jd clustering algorithm model, the mean spring accuracy within each cluster is calculated over the temperature range.

[0120] It should be noted that the trained temperature-spring accuracy Jd clustering algorithm model, within the cluster, uses the temperature range covered by the cluster as a limiting condition, and the average accuracy calculated is a representative data feature within the cluster.

[0121] Obtain the average precision of all clusters, sort the average precision in descending order, and obtain the highest precision;

[0122] Find the temperature range of the cluster with the highest accuracy, and then obtain the optimal temperature range;

[0123] The optimal temperature is obtained by summing the optimal temperature ranges and taking the average value.

[0124] For example, cluster 1: temperature range [700℃, 710℃], average accuracy 85%; cluster 2: temperature range [711℃, 715℃], average accuracy 95%; cluster 3: temperature range [716, 720℃], average accuracy 90%; the optimal temperature range is [711℃, 715℃]; and the optimal temperature is 713℃.

[0125] The technical solution of this embodiment is as follows: based on the degree analysis signal, obtain the spring accuracy Jd corresponding to different temperatures, use a clustering algorithm to establish a clustering algorithm model of temperature-spring accuracy Jd, and calculate the optimal temperature.

[0126] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A method for preparing a high-precision high-temperature alloy spring, characterized in that: Preparation methods include: S1. Nickel-based alloys are selected as raw materials, and aluminum, molybdenum, and niobium elements are added to the materials; S2. Perform three heat treatments on the alloy; S3. Perform wire drawing on the heat-treated alloy; S4. Use a CNC spring forming machine to wind the wire into shape; S5. Polish the spring surface to remove residual oxides from the processing, and plate chromium onto the spring surface as a high-temperature resistant protective coating. In S5, it includes: Step 1: Obtain the spring length data, calculate the spring length difference with the target spring length value, mark springs whose spring length difference does not meet the threshold as abnormal springs, calculate the abnormal length ratio Cd by the average length difference of abnormal springs and the spring length difference threshold, calculate the proportion of abnormal springs to obtain the abnormal spring number ratio SI, calculate the spring accuracy Jd based on the abnormal spring number ratio SI and the abnormal length ratio Cd, and generate an anomaly analysis signal if the spring accuracy Jd ≥ the spring accuracy threshold Jdy. The spring accuracy Jd is obtained as follows: The spring accuracy Jd is calculated based on the abnormal spring quantity ratio SI and the abnormal length ratio Cd. Through the formula: , obtain the spring accuracy Jd, where a and b are preset proportional coefficients; The abnormal analysis signal is generated in the following way: Compare the spring accuracy Jd with the spring accuracy threshold Jdy; If the spring accuracy Jd ≥ the spring accuracy threshold Jdy, an anomaly analysis signal is generated. Step 2: Based on the anomaly analysis signal, collect data from multiple production batches to obtain the change in spring accuracy Jd under different temperature conditions, plot the temperature-spring accuracy Jd curve, plot a reference line using the spring accuracy threshold Jdy as a reference value, calculate the abnormal curve segment where the spring accuracy Jd exceeds the spring accuracy threshold Jdy, obtain the abnormal area enclosed by the abnormal curve segment and the reference line, obtain the temperature block length value corresponding to the abnormal curve segment, calculate the average abnormal area ratio Yc and the temperature block length ratio Wd, and calculate the anomaly degree coefficient Cx. The anomaly degree coefficient Cx ≥ the anomaly degree coefficient threshold Cxz, and generate the degree analysis signal. The anomaly degree coefficient Cx is obtained as follows: The anomaly severity coefficient Cx is calculated based on the mean anomaly area ratio Yc and the temperature block length ratio Wd. Through the formula: The anomaly degree coefficient Cx was calculated. The average abnormal area ratio Yc is obtained as follows: Plot the temperature-spring accuracy Jd curve in a two-dimensional rectangular coordinate system with temperature as the X-axis and spring accuracy Jd as the Y-axis. Using the spring accuracy threshold Jdy as a reference value, draw a horizontal reference line in the coordinate system; Obtain the curve segment where the spring accuracy Jd exceeds the spring accuracy threshold Jdy; this is the abnormal curve segment. The area enclosed by a single abnormal curve segment and the reference line is used to obtain the abnormal area. Obtain the area enclosed by a single abnormal curve segment and the X-axis to get the curve area; The ratio of the abnormal area to the curve area is calculated to obtain the abnormal area ratio. Obtain the ratio of abnormal area for all abnormal curve segments and sum them to get the total ratio of abnormal area; The ratio of the total abnormal area to the number of abnormal curve segments is processed to obtain the average abnormal area ratio, and the average abnormal area ratio is marked as Yc. Obtain the abnormal temperature segment corresponding to the abnormal curve segment on the X-axis, and calculate the length of the abnormal temperature segment; The ratio of the abnormal temperature segment length to the total temperature segment length is calculated to obtain the temperature block length ratio, which is then denoted as Wd. Step 3: Based on the degree analysis signal, obtain the spring accuracy Jd corresponding to different temperatures, use a clustering algorithm to establish a clustering algorithm model of temperature-spring accuracy Jd, and calculate the optimal temperature.

2. The method for preparing a high-precision high-temperature alloy spring according to claim 1, characterized in that: The method for obtaining the number ratio SI of the abnormal springs is as follows: Obtain the length data of the spring sample, and subtract the spring length of the spring sample from the target spring length value to obtain the spring length difference; Compare the spring length difference with the spring length difference threshold; Springs whose spring length difference does not fall within the threshold range are marked as abnormal springs; Obtain the number of abnormal springs, and calculate the ratio of the number of abnormal springs to the total number of springs tested to obtain the abnormal spring count ratio. Mark the abnormal spring count ratio as Sl.

3. The method for preparing a high-precision high-temperature alloy spring according to claim 2, characterized in that: The abnormal length ratio Cd is obtained as follows: The critical spring length difference is obtained by subtracting the length difference of the abnormal spring from the endpoint of the nearest adjacent spring length difference threshold range and taking the absolute value. Obtain the critical length difference of all abnormal springs, sum the critical length differences of all abnormal springs, and get the total critical length difference value. The average critical difference of spring length is obtained by comparing the total difference in the critical length of the springs with the total number of all abnormal springs. The abnormal length ratio is obtained by comparing the average critical difference of spring length with the length of the spring length difference threshold range. The abnormal length ratio is then labeled as Cd.

4. The method for preparing a high-precision high-temperature alloy spring according to claim 1, characterized in that: The method for generating the degree analysis signal is as follows: Compare the anomaly severity coefficient Cx with the anomaly severity coefficient threshold Cxz; If the anomaly degree coefficient Cx ≥ the anomaly degree coefficient threshold Cxz, a degree analysis signal is generated.

5. The method for preparing a high-precision high-temperature alloy spring according to claim 1, characterized in that: The optimal temperature is obtained as follows: Based on the trained temperature-spring accuracy Jd clustering algorithm model, the mean spring accuracy within each cluster is calculated over the temperature range. Obtain the average precision of all clusters, sort the average precision in descending order, and obtain the highest precision; Find the temperature range of the cluster with the highest accuracy, and then obtain the optimal temperature range; The optimal temperature is obtained by summing and averaging the optimal temperature ranges.

6. The method for preparing a high-precision high-temperature alloy spring according to claim 5, characterized in that: The temperature-spring accuracy Jd clustering algorithm model is constructed as follows: SS1. Perform data preprocessing on the acquired temperature-spring accuracy Jd set H, removing outliers and standardizing the data; SS2. Select the K-Means clustering algorithm and use the elbow rule to determine the number K; SS3: Randomly initialize three cluster centers, calculate the distance between the data point and the three cluster centers, treat the data pair as the data point within the cluster center, and assign the data point to the cluster containing the nearest cluster center; SS4. Recalculate the center of each cluster, which is the mean of all data points in the cluster. Repeat step SS3 until the cluster centers no longer change or the predetermined number of iterations is reached.

Images