Digital production method and system for liquid accumulator of air conditioner compressor

Abstract

The embodiment of the invention discloses a digital production method and system for a liquid accumulator of an air conditioner compressor, belongs to the technical field of spinning prediction, and solves the problem of low production efficiency caused by untimely spinning regulation in the prior art. A first starting instruction of the spinning machine is received, a product code is generated, and spinning treatment and stamping shaping treatment are conducted on the air outlet end of the barrel blank through the spinning machine; receiving a spinning data acquisition processing instruction, and performing fitting adjustment on the acquired data; spinning machine operation data are obtained, and spinning machine operation fitting curves corresponding to different operation conditions are generated based on the spinning machine operation data; the spinning machine is regulated and controlled in advance based on the adjusted actual fitting curve and the spinning machine operation fitting curve; a spinning grooving instruction and a spinning machine second starting instruction are received, and the remaining part of the liquid storage device is produced; and obtaining production information corresponding to different production links of the liquid storage device, and mapping the production information and the product code to trace the production process of the liquid storage device.

Description

Technical Field

[0001] The present application relates to the field of spinning prediction technology, and in particular to a digital production method and system for an air-conditioning compressor accumulator. Background Art

[0002] The liquid receiver is an important component in the refrigeration system. Its main function is to store liquid refrigerant and ensure that the moisture in the refrigerant is effectively extracted. In the field of liquid receiver manufacturing, in order to improve the sealing performance of the liquid receiver, optimize fluid flow, enhance structural strength, improve production efficiency and meet specific design requirements, the spinning process of the liquid receiver end plays a vital role.

[0003] In the prior art, since many factors are involved in the spinning process, and the interactions between different factors are complex and difficult to accurately quantify. Therefore, even experienced technicians find it difficult to accurately predict the accuracy after spinning based on experience alone. In addition, fluctuations in spinning accuracy will also affect the overall performance of the reservoir. As an important component in the fluid system, the spinning accuracy of the reservoir's end is directly related to the sealing, stability and durability of the fluid. If the spinning accuracy is not up to standard, it may lead to problems such as fluid leakage and pressure fluctuations, which will affect the normal operation of the entire fluid system.

[0004] The existing liquid reservoir spinning technology often finds that the accuracy does not meet the standard during the inspection stage after spinning, and then adjusts the key parameters such as spinning angle and pressure, resulting in high production costs and low production efficiency. Summary of the invention

[0005] The embodiments of the present application provide a digital production method and system for air-conditioning compressor liquid reservoirs, which are used to solve the following technical problems: the existing liquid reservoir spinning technology often finds that the accuracy does not meet the standards during the inspection stage after spinning, and then adjusts key parameters such as the spinning angle and pressure, resulting in high production costs and low production efficiency.

[0006] The present application embodiment adopts the following technical solutions:

[0007] The embodiment of the present application provides a digital production method for an air-conditioning compressor liquid reservoir, including obtaining liquid reservoir preparation completion information and generating a product code; wherein the liquid reservoir preparation includes a barrel blank, an air inlet straight pipe, an air outlet elbow, a middle partition and a filter; receiving a first start-up instruction of a spinning machine, and performing spinning and stamping shaping on the outlet end of the barrel blank through the spinning machine; receiving a spinning data collection and processing instruction, dividing the collected radian angle data of the spinning closing of the end of the liquid reservoir into intervals, fitting the radian angle data based on the radian angle data difference corresponding to different intervals, obtaining an actual fitting curve, and based on the shell and tube of the liquid reservoir data, adjust the actual fitting curve; obtain the operating data of the spinning machine, and generate the spinning machine operating fitting curves corresponding to different operating conditions based on the spinning machine operating data; pre-regulate the spinning machine based on the adjusted actual fitting curve and the spinning machine operating fitting curve; when the structural data of the air outlet meets the first preset structural data, receive the spinning grooving instruction and the second starting instruction of the spinning machine, and produce the remaining part of the liquid reservoir; obtain the production information corresponding to different production links of the liquid reservoir, map the production information corresponding to different production links with the product code, so as to trace the production process of the liquid reservoir through the mapping relationship.

[0008] The embodiment of the present application uses interval division and fitting technology to extract continuous and smooth curves from discrete data points, which more intuitively reflects the trend of the arc angle change of the spinning closure. Secondly, the embodiment of the present application introduces tube shell data to make the adjusted fitting curve closer to the actual production situation, reduce the prediction error caused by tube shell differences, and improve the accuracy of future spinning predictions. In addition, the embodiment of the present application generates fitting curves under different operating conditions based on the operating data of the spinning machine, which can fully reflect the performance characteristics of the spinning machine. Through prediction, the spinning process can be adjusted before production to avoid errors and waste in the production process and improve production efficiency. The embodiment of the present application stores all production information of the reservoir production process through product coding, and can trace the production information, thereby providing more comprehensive production monitoring and management capabilities.

[0009] In one implementation of the present application, the production information corresponding to different production links are mapped with the product code to trace the production process of the liquid reservoir through the mapping relationship, specifically including: uploading the production information to the blockchain; dividing the production information into ordinary information and sensitive information based on the data type of the production information, and homomorphically encrypting the sensitive information to obtain encrypted sensitive information; performing hash calculations on the encrypted sensitive information and the ordinary information to generate a hash value uniquely corresponding to the production information of the current production link; based on the production information and the hash value, generating processing information codes corresponding to different production links of the liquid reservoir; based on the production sequence, constructing a connection network between adjacent production links, and based on the connection network, associating the processing information codes corresponding to different production links to construct a relationship association network corresponding to the processing information codes; mapping the relationship association network with the product code to trace the liquid reservoir product through the mapping relationship.

[0010] In one implementation of the present application, the collected radian angle data of the spun closing at the end of the liquid reservoir are divided into intervals, and the radian angle data are fitted based on the differences in the radian angle data corresponding to different intervals to obtain an actual fitting curve, which specifically includes: obtaining the radian angle data of the spun closing by means of angle sensors arranged at different positions on the end of the liquid reservoir; dividing the radian angle data of the spun closing into multiple first intervals according to the shape of the spun closing at the end of the liquid reservoir; obtaining adjacent change differences of the radian angle data in the first interval, determining interval connection points based on the adjacent change differences, and dividing the first interval into multiple second intervals based on the interval connection points; determining interval radian differences corresponding to the multiple second intervals, and matching the multiple second intervals with polynomials of different orders based on the interval radian differences; fitting based on polynomials of different orders corresponding to the multiple second intervals and the radian angle data of the spun closing corresponding to the multiple second intervals to obtain an actual fitting curve.

[0011] In one implementation of the present application, the actual fitting curve is adjusted based on the tube shell data of the liquid reservoir, specifically including: obtaining the tube shell thickness in the tube shell data, and determining a first adjustment coefficient based on the difference between the tube shell thickness and the reference tube shell thickness; constructing a clamping error detection model based on the relationship between the historical tube shell clamping vertical angle in the historical tube shell data and the historical fitting curve adjustment coefficient; obtaining the tube shell clamping vertical angle in the tube shell data, inputting the tube shell clamping vertical angle into the clamping error detection model to obtain a second adjustment coefficient; and adjusting the actual fitting curve based on the first adjustment coefficient and the second adjustment coefficient.

[0012] In one implementation of the present application, the operation data of the spinning machine is obtained, and the operation fitting curves of the spinning machine corresponding to different operating conditions are generated based on the operation data of the spinning machine, specifically including: obtaining the operation data of the spinning machine; wherein the operation data includes at least temperature, pressure and rotation speed; fitting the operation data by statistical methods to obtain the fitting curve groups corresponding to different operating conditions; normalizing the horizontal coordinates of the operation data of the spinning machine and the fitting curve group; determining a reference curve in the fitting curve group, comparing other curves in the fitting curve group with the reference curve, and determining the relative value of the vertical coordinate corresponding to the same reference point as the reference curve in the other curves; fitting based on the relative value of the vertical coordinate and the normalized data to obtain the spinning machine operation fitting curves corresponding to the spinning machine under different operating conditions.

[0013] In one implementation of the present application, the spinning machine is controlled in advance based on the adjusted actual fitting curve and the spinning machine operation fitting curve, specifically including: extracting key parameters of the adjusted actual fitting curve; wherein the extracted features include at least one of the curve slope, curve curvature and curve extreme points; inputting the extracted key parameters and the corresponding spinning machine operation fitting curve under different operating conditions into the spinning prediction model, so as to output the predicted fitting curve corresponding to the future time period through the spinning prediction model; determining the first contribution value of different operating data features to the predicted fitting curve corresponding to the future time period through the SHAP algorithm, and drawing a feature global interpretation map based on the contribution value; constructing a local linear model through the LIME algorithm, determining the second contribution value of each different operating data feature to the predicted fitting curve corresponding to the future time period through the weight of the local linear model, and drawing a feature local interpretation map based on the second contribution value; constructing a future spinning prediction fitting curve based on the feature global interpretation map and the feature local interpretation map, and comparing the future spinning prediction fitting curve with the target curve, and controlling the spinning machine in advance based on the comparison result.

[0014] In one implementation of the present application, the future spinning prediction fitting curve is compared with the target curve, and the spinning of the reservoir end is controlled in advance based on the comparison result, specifically including: comparing the future spinning prediction fitting curve with the target curve to determine the error point between the two, and dividing the future spinning prediction fitting curve into multiple sections based on the error point; determining the error type based on the error values ​​corresponding to the multiple sections; wherein the error type includes position error, shape error and size error; inputting the error type and the error value into a preset error control model, and outputting the corresponding spinning machine control strategy through the preset error control model; obtaining the spinning processing error threshold of the reservoir end, comparing the data in the section with the spinning processing error threshold, and determining the threshold reference point; determining the error adjustment duration based on the distance between the threshold reference point and the initial point of the section, and determining the control time point before the initial point of the section based on the adjustment duration, so as to implement the spinning machine control strategy at the control time point.

[0015] In one implementation of the present application, the error type is determined based on the error values ​​corresponding to multiple segments, specifically including: generating an error value fluctuation sequence corresponding to each segment based on the error value data within the segment; determining the regional extreme value points in the error value fluctuation sequence, and determining the minimum value in the sequence after the regional extreme value points, and constructing an error detection sequence based on the minimum value; detecting the regional extreme value points based on the error detection sequence, and constructing a reference sequence based on the regional extreme value points that meet the detection conditions; determining the feature points of the reference sequence, and determining the sequence floating pattern of the reference sequence based on the feature points; constructing a sequence floating pattern set based on the sequence floating patterns corresponding to multiple segments, and matching the sequence floating pattern set with a preset error type library to determine the error type; wherein the preset error type library includes multiple floating pattern sets, and also includes error types corresponding to the multiple floating pattern sets.

[0016] In one implementation of the present application, when the structural data of the air outlet end meets the first preset structural data, a spinning grooving instruction and a second starting instruction of the spinning machine are received to produce the remaining part of the liquid reservoir, specifically including: welding the air outlet elbow to the air outlet end when the structural data of the air outlet end meets the first preset structural data; installing the middle partition plate to the first preset position along the axial direction of the cylinder blank; receiving the first spinning machine grooving instruction, and grooved the outer wall of the cylinder blank corresponding to the first preset position by the grooving machine, so that the outer wall of the cylinder blank convexes radially toward the inside of the cylinder along the cylinder. The filter screen is installed to a second preset position along the axial direction of the cylinder blank; a second grooving instruction of the spinning machine is received, and a groove is made on the outer wall of the cylinder blank corresponding to the second preset position by the grooving machine, so that the outer wall of the cylinder blank protrudes toward the inside of the cylinder along the radial direction of the cylinder, and the filter screen is limited axially; a second start-up instruction of the spinning machine is received, and the air inlet end of the cylinder blank is spun by the spinning machine, and when the structural data of the air inlet end meets the second preset structural data, the air inlet straight pipe is welded to the air inlet end, so as to realize the digital production of the air conditioner compressor liquid reservoir.

[0017] The embodiment of the present application provides a digital production system for the liquid reservoir of an air-conditioning compressor, characterized in that the system includes a liquid reservoir preparation unit, a spinning unit, a spinning data acquisition and processing unit, a remaining part production unit and a traceability unit; the liquid reservoir preparation unit is used to obtain the completion information of the liquid reservoir preparation and generate a product code; wherein the liquid reservoir preparation includes a barrel blank, an air inlet straight pipe, an air outlet bent pipe, a middle partition and a filter; the spinning unit is used to receive a first start-up instruction of a spinning machine to start the spinning machine to perform spinning and stamping shaping on the air outlet end of the barrel blank; the spinning data acquisition and processing unit is used to receive a spinning data acquisition and processing instruction, divide the collected radian angle data of the spinning closure of the liquid reservoir end into intervals, and divide the radian angle data into intervals based on the difference in radian angle data corresponding to different intervals. The spinning data acquisition and processing unit is further used to obtain the operation data of the spinning machine, generate the operation fitting curves of the spinning machine corresponding to different operation conditions based on the operation data of the spinning machine, and adjust the spinning machine in advance based on the adjusted actual fitting curve and the operation fitting curve of the spinning machine; the remaining part production unit is used to receive the spinning grooving instruction and the second starting instruction of the spinning machine when the structural data of the air outlet end meets the first preset structural data, and produce the remaining part of the liquid reservoir; the tracing unit is used to obtain the production information corresponding to different production links of the liquid reservoir, map the production information corresponding to different production links with the product code, so as to trace the production process of the liquid reservoir through the mapping relationship.

[0018] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects: The embodiments of the present application extract continuous and smooth curves from discrete data points through interval division and fitting technology, which more intuitively reflects the trend of the curvature angle change of the spinning closure. Secondly, the embodiment of the present application introduces tube shell data to make the adjusted fitting curve closer to the actual production situation, reduce the prediction error caused by tube shell differences, and improve the accuracy of future spinning predictions. In addition, the embodiment of the present application generates fitting curves under different operating conditions based on the operating data of the spinning machine, which can fully reflect the performance characteristics of the spinning machine. Through prediction, the spinning process can be adjusted before production to avoid errors and waste in the production process and improve production efficiency. The embodiment of the present application stores all production information of the reservoir production process through product coding, and can trace the production information, thereby providing more comprehensive production monitoring and management capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. In the drawings:

[0020] Figure 1 A flow chart of a digital production method for an air-conditioning compressor accumulator provided in an embodiment of the present application;

[0021] Figure 2 A flow chart of a liquid reservoir processing provided in an embodiment of the present application;

[0022] Figure 3 A schematic structural diagram of a digital production system for an air-conditioning compressor fluid reservoir provided in an embodiment of the present application.

[0023] Reference numerals:

[0024] 1: cylinder, 11: outlet end, 111: outlet elbow installation hole, 112: flat part, 113: arc part, 12: inlet end, 121: inlet straight pipe installation hole, 21: outlet elbow, 22: inlet straight pipe, 23: welding flange, 231: first limiting step, 232: second limiting step; 3: filter screen; 4: middle partition;

[0025] 200: Digital production system for air-conditioning compressor accumulator, 201: accumulator material preparation unit, 202: spinning unit, 203: spinning data acquisition and processing unit, 204: remaining part production unit, 205: traceability unit. DETAILED DESCRIPTION

[0026] The embodiments of the present application provide a digital production method and system for an air-conditioning compressor accumulator.

[0027] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0028] The technical solution proposed in the embodiment of the present invention is described in detail below with reference to the accompanying drawings.

[0029] Figure 1 A flow chart of a digital production method for an air-conditioning compressor accumulator provided in an embodiment of the present application, such as Figure 1 As shown, the method for predicting and controlling the spinning pressure at the end of the liquid reservoir includes the following steps:

[0030] Step 101: Obtain the liquid storage device preparation completion information and generate a product code.

[0031] In one embodiment of the present application, the liquid reservoir material preparation includes a barrel blank, an air inlet straight pipe, an air outlet elbow, a middle baffle and a filter. Other components to be installed are also included. The barrel blank is cut from a fixed length of pipe according to the specifications of the liquid reservoir barrel. The system determines that all materials for producing the liquid reservoir have been prepared by receiving the liquid reservoir material preparation completion information.

[0032] In one embodiment of the present application, a QR code label is attached to the liquid reservoir product to be produced, and the liquid reservoir identity data information is given. The product information corresponding to each process in the production process, production photos and videos, important process management information records, and visual inspection information and other data are connected to the QR code label, so that the entire production process of the liquid reservoir can be traced through the QR code label.

[0033] Step 102, receiving a first start instruction of a spinning machine, and performing spinning and stamping shaping on the outlet end of the barrel blank through the spinning machine.

[0034] In one embodiment of the present application, a cut cylindrical blank is installed on the pressure spindle of the spinning machine. After receiving the first start-up command of the spinning machine, the cylindrical blank is driven to rotate by the pressure spindle of the spinning machine and cooperates with the spinning wheel to spin one end of the cylindrical blank into a spherical shape, making it an initial air outlet end. In the embodiment of the present application, the pressure spindle drives the cylindrical blank to move and contact with the bottom mold, drives the upper mold to move, cooperates with the bottom mold to shape the initial air outlet end so that the bottom of the initial air outlet end is flat, and punches out the air outlet elbow installation hole for installing the air outlet elbow in the center of the initial air outlet end, completing the processing of the air outlet end.

[0035] Step 103, receiving a spinning data collection and processing instruction, dividing the collected radian angle data of the spinning closure at the end of the liquid reservoir into intervals, fitting the radian angle data based on the radian angle data differences corresponding to different intervals, obtaining an actual fitting curve, and adjusting the actual fitting curve based on the tube shell data of the liquid reservoir.

[0036] In one embodiment of the present application, the radian angle data of the spun closure is obtained by means of angle sensors arranged at different positions at the end of the reservoir. According to the shape of the spun closure at the end of the reservoir, the radian angle data of the spun closure is divided into a plurality of first intervals. The adjacent change difference of the radian angle data in the first interval is obtained, and based on the adjacent change difference, the interval connection point is determined, and the first interval is divided into a plurality of second intervals based on the interval connection point. The interval radian difference corresponding to the plurality of second intervals is determined, and polynomials of different orders are matched to the plurality of second intervals based on the interval radian difference. Based on the polynomials of different orders corresponding to the plurality of second intervals, and the radian angle data of the spun closure corresponding to the plurality of second intervals, fitting is performed to obtain an actual fitting curve.

[0037] Specifically, during the process of the spinning machine processing the outlet end, angle sensors are set at different positions of the end of the liquid reservoir, and the radian angle data at the spinning closing is collected through the sensors. For example, angle sensors are respectively set at the upper, middle and lower positions of the end of the liquid reservoir. According to the shape of the spinning closing of the end of the liquid reservoir, the collected radian angle data is divided into multiple first intervals.

[0038] Specifically, based on the acquired arc angle data of the spun joint, the shape characteristics of the spun joint are determined, and based on the shape characteristics of the spun joint, the number of preset first intervals is determined. Based on the arc angle data of the spun joint, a data density distribution diagram is drawn. The data density distribution diagram is tested to determine the density value between different data. The density value is compared with the preset area division density threshold, and a reference point corresponding to the preset area division density threshold is determined on the density distribution diagram. The distance between the reference points is determined, and the width of different first intervals is determined based on the distance.

[0039] Specifically, a set of arc angle data of the spinning closure is obtained, and according to the distribution trend of the data, the statistics are calculated. For example, if the data shows a trend of first increasing and then decreasing, and there is an obvious inflection point. At this time, three preset first intervals can be preliminarily set: the increasing interval before the inflection point, the turning interval at the inflection point, and the decreasing interval after the inflection point. The above spinning closure data is converted into a density curve, which will have an obvious peak at the inflection point and is relatively smooth in the increasing and decreasing intervals.

[0040] Furthermore, the embodiment of the present application determines the peak points in the data density distribution diagram through a density peak detection algorithm, and compares these peak points with the preset area division density threshold to determine the reference point corresponding to the threshold. For example, in the density distribution diagram, the peak at the inflection point is determined as the peak point, the density threshold is set based on a preset ratio of the peak height, and the reference point corresponding to the threshold is found. Furthermore, the difference between adjacent reference points is calculated, and through these differences, the width of different first intervals can be determined more accurately, thereby achieving a fine division of data distribution.

[0041] Furthermore, in each first interval, adjacent change differences of the radian angle data are obtained, and interval connection points are determined based on these differences. Then, the first interval is further divided into multiple second intervals based on these connection points, so as to more finely analyze the data changes in each interval. For example, in the second first interval [30°, 45°], if the change difference of adjacent data points is large, for example, due to the uneven spinning, adjacent data with large differences appear in the change from 30° to 45°, and the rate of change from 30° to 35° is different from the rate of change from 35° to 45°, the interval connection point 35° is determined based on these differences, thereby dividing this interval into two second intervals [30°, 35°] and [35°, 45°].

[0042] Further, for each second interval, determine the corresponding interval radian difference, that is, the difference between the maximum radian angle and the minimum radian angle in the interval. Then, based on these interval radian differences, polynomials of different orders are matched for each second interval. For the second second interval [30°, 35°], if its interval radian difference is 5°, it can be matched with a first-order or second-order polynomial. If the data changes smoothly, a first-order polynomial is matched; if the data changes more complex, a high-order polynomial is matched. The embodiment of the present application is provided with a polynomial matching data table, which includes different interval radian differences and polynomials corresponding to different radian differences. Based on the polynomials of different orders corresponding to each second interval and the radian angle data in these intervals, fitting is performed to obtain an actual fitting curve. The fitted curve can reflect the radian angle change trend of the spinning closure at the end of the reservoir.

[0043] In one embodiment of the present application, a tube shell thickness in the tube shell data is obtained, and a first adjustment coefficient is determined based on a difference between the tube shell thickness and a reference tube shell thickness. A clamping error detection model is constructed based on a relationship between a historical tube shell clamping vertical angle in the historical tube shell data and an adjustment coefficient of a historical fitting curve. The tube shell clamping vertical angle in the tube shell data is obtained, and the tube shell clamping vertical angle is input into the clamping error detection model to obtain a second adjustment coefficient. Based on the first adjustment coefficient and the second adjustment coefficient, the actual fitting curve is adjusted.

[0044] Specifically, the tube shell thickness is extracted from the tube shell data, the tube shell thickness is compared with the reference tube shell thickness, the difference is calculated, and the first adjustment coefficient is determined according to the difference, which reflects the influence of the tube shell thickness on the fitting curve. For example, the tube shell thickness extracted from the database is 2.6mm, and the difference between the tube shell thickness and the reference thickness is calculated, that is, 2.5mm-2.0mm=0.6mm. According to the preset difference adjustment rule, the adjustment coefficient increases by 0.05 for every 0.1mm increase in thickness. Therefore, for a difference of 0.5mm, the first adjustment coefficient is 0.6 / 0.1×0.05=0.3.

[0045] The clamping error detection model is constructed by using the relationship between the historical tube shell clamping vertical angle and the historical fitting curve adjustment coefficient in the historical tube shell data. The clamping error detection model can predict the clamping error according to the clamping vertical angle. The training process of the model is to use the historical tube shell clamping vertical angle as the input sample, and the historical fitting curve adjustment coefficient corresponding to the input sample as the output sample, and train the preset neural network model to obtain the clamping error detection model. The tube shell clamping vertical angle is extracted from the current tube shell data, and the angle is input into the clamping error detection model to obtain the predicted second adjustment coefficient, which reflects the influence of the clamping error on the fitting curve. For example, the clamping vertical angle extracted from the current tube shell data is 87°, and the clamping vertical angle is input into the clamping error detection model to obtain a clamping error of 0.03. The adjustment rule is set according to the predicted clamping error. For example, for every increase of 0.01 in the clamping error, the adjustment coefficient is reduced by 0.02. Therefore, for a clamping error of 0.03, the second adjustment coefficient is 0.03 / 0.01×(-0.02)=-0.06.

[0046] Furthermore, the actual fitting curve is adjusted by combining the first adjustment coefficient and the second adjustment coefficient, and the purpose of the adjustment is to make the fitting curve more consistent with the actual shape and clamping condition of the tube shell. The embodiment of the present application adopts a weighted summation method to combine the adjustment coefficients, and the weights are equal (i.e., each accounts for 50%), so the combined adjustment coefficient is (0.3-0.06) / 2=0.12.

[0047] Step 104: Acquire the spinning machine operation data, and generate spinning machine operation fitting curves corresponding to different operation conditions based on the spinning machine operation data.

[0048] In one embodiment of the present application, the operating data of the spinning machine is obtained; wherein the operating data includes at least temperature, pressure and rotation speed. The operating data is fitted by a statistical method to obtain a group of fitting curves corresponding to different operating conditions. The horizontal coordinates of the operating data of the spinning machine and the fitting curve group are normalized. A reference curve is determined in the fitting curve group, and other curves in the fitting curve group are compared with the reference curve, and the relative value of the ordinate corresponding to the same reference point as the reference curve is determined in the other curves. Based on the relative value of the ordinate and the normalized data, a fitting curve of the spinning machine operation corresponding to the spinning machine under different operating conditions is obtained.

[0049] Specifically, key parameters of the spinning machine during operation are obtained, including temperature, pressure and rotation speed. These data can be directly measured by sensors and stored in a data recording system. These data are fitted using statistical methods to obtain a corresponding set of fitting curves under different operating conditions (such as different combinations of temperature, pressure and rotation speed). Each curve represents the data trend under a specific combination of operating conditions. For example, for each temperature, a curve about pressure and rotation speed is fitted, and for each pressure, a curve about temperature and rotation speed is fitted, and so on. According to the series of fitting curves obtained, a fitting curve group is formed.

[0050] Furthermore, the horizontal coordinates (such as pressure, temperature or rotation speed) of the operating data of the spinning machine and the fitting curve group are normalized. A reference curve is determined in the fitting curve group, and each curve in the fitting curve group is compared with the reference curve to find out their relative values ​​of the vertical coordinates (such as temperature or rotation speed) at the same reference point (such as the same normalized pressure value). For example, for a point where the normalized pressure value is 0.5, the temperature values ​​of the reference curve and other curves at this point are compared to obtain relative values, which reflect the relative changes in the performance of the spinning machine under different operating conditions. The calculated relative values ​​and the normalized data are used as new input data, and the new data are fitted using statistical methods to obtain a spinning machine operation fitting curve that describes the performance changes of the spinning machine under different operating conditions.

[0051] Step 105: pre-regulate the spinning machine based on the adjusted actual fitting curve and the spinning machine operation fitting curve.

[0052] In one embodiment of the present application, the key parameters of the adjusted actual fitting curve are extracted; wherein the extracted features include at least one of the slope of the curve, the curvature of the curve, and the extreme point of the curve. The extracted key parameters and the corresponding spinning machine operation fitting curve under different operating conditions are input into the spinning prediction model to output the prediction fitting curve corresponding to the future time period through the spinning prediction model. The first contribution value of different operating data features to the prediction fitting curve corresponding to the future time period is determined by the SHAP algorithm, and a feature global interpretation diagram is drawn based on the contribution value. A local linear model is constructed by the LIME algorithm, and the second contribution value of each different operating data feature to the prediction fitting curve corresponding to the future time period is determined by the weight of the local linear model, and a feature local interpretation diagram is drawn based on the second contribution value. Based on the feature global interpretation diagram and the feature local interpretation diagram, a future spinning prediction fitting curve is constructed, and the future spinning prediction fitting curve is compared with the target curve, and the spinning machine is adjusted in advance based on the comparison result.

[0053] Specifically, the key parameters of the adjusted actual fitting curve are extracted, and the extracted key parameters and the corresponding spinning machine operation fitting curve under different operating conditions are used as input features and input into the spinning prediction model, and the prediction fitting curve corresponding to the future time period is output through the spinning prediction model. The training process of the spinning prediction model is as follows: the key parameters of the historical fitting curve and the fitting curve corresponding to different historical operating conditions are used as input samples, and the historical fitting curve corresponding to the input sample is used as the output sample, and the preset neural network model is trained to obtain the spinning prediction model.

[0054] Furthermore, the SHAP algorithm is used to determine the first contribution value of different operating data features to the prediction fitting curve corresponding to the future time period, that is, the importance of each feature to the prediction result. Then, a feature global interpretation diagram is drawn based on these contribution values ​​to intuitively show the overall impact of each feature on the prediction result. For example, the prediction results of the model are interpreted by the SHAP algorithm, and it is determined that the pressure feature has the largest contribution value to the prediction result, followed by the temperature feature. These contribution values ​​are visualized as a global interpretation diagram, where the length of each feature represents the size of its contribution value, and the color represents the positive or negative contribution value.

[0055] Furthermore, the LIME algorithm is used to construct a local linear model, and the second contribution value of each different operating data feature to the prediction fitting curve corresponding to the future time period is determined. Based on these contribution values, a feature local interpretation diagram is drawn to intuitively show the local impact of each feature on the prediction result near a specific data point. For example, the prediction results of the model under specific operating conditions are explained by the LIME algorithm, and it is found that under this condition, the local contribution value of the temperature feature to the prediction result is the largest. These local contribution values ​​are visualized as a local interpretation diagram, where the length of each feature represents the size of its contribution value near a specific data point, and the color represents the positive or negative contribution value.

[0056] Furthermore, the predicted numerical interpretation corresponding to the future spinning prediction fitting curve is determined based on the feature global interpretation diagram and the feature local interpretation diagram. The global interpretation diagram provides the overall impact of each feature on the prediction result, while the local interpretation diagram provides the local impact of each feature near a specific data point. For example, the pressure feature in the global interpretation diagram has the largest contribution to the prediction result, while the temperature feature in the local interpretation diagram has the largest contribution to the prediction result under specific operating conditions. This means that pressure is a key factor affecting the performance of the spinning machine in general, but the influence of temperature is more significant under specific conditions. Therefore, it can be proposed that when optimizing the performance of the spinning machine, priority should be given to adjusting the pressure parameters, and special attention should be paid to the temperature parameters under specific conditions.

[0057] Furthermore, if the global interpretation diagram shows that a certain feature has a great impact on the prediction results, the weight of the feature can be increased in the spinning prediction model; if the local interpretation diagram shows that a certain feature has a significant impact under certain specific conditions, conditional judgment can be introduced into the spinning prediction model to dynamically adjust the weight of the feature. The optimized spinning prediction model will be able to more accurately reflect the performance change trend of the spinning machine in the future time period, and the operating data will be input into the optimized model to generate the future spinning prediction fitting curve.

[0058] In one embodiment of the present application, the future spinning prediction fitting curve is compared with the target curve to determine the error point between the two, and based on the error point, the future spinning prediction fitting curve is divided into multiple sections. Based on the error values ​​corresponding to the multiple sections, the error type is determined; wherein the error type includes position error, shape error and size error. The error type and the error value are input into the preset error control model, and the corresponding spinning machine control strategy is output through the preset error control model. The spinning processing error threshold of the reservoir end is obtained, and the data in the section is compared with the spinning processing error threshold to determine the threshold reference point. Based on the distance between the threshold reference point and the initial point of the section, the error adjustment duration is determined, and the control time point is determined before the initial point of the section based on the adjustment duration, so as to implement the spinning machine control strategy at the control time point.

[0059] Specifically, the future spinning prediction fitting curve is compared with the target curve to determine the difference between them, that is, the error point. Based on these error points, the predicted fitting curve is divided into multiple sections. Among them, each section represents a different error feature. Then, according to the error value of each section, the type of error is determined, such as position error, the position offset of the curve on the X-axis; shape error, the shape of the curve does not match the target; size error, the size deviation of the curve on the Y-axis. The determined error type and error value are input into the preset error control model. The preset error control model outputs the corresponding spinning machine control strategy based on the input error information, such as adjusting the spinning speed, pressure or path. For example, if the model detects a position error, it will output a suggestion to increase or decrease the movement of the spinning head to adjust the position. The training process of the preset error control model is to take the historical error type and historical error value as input samples, and the control strategy corresponding to the input sample as the output sample, and train the preset neural network to obtain the preset error control model.

[0060] Furthermore, the embodiment of the present application sets a spinning error threshold, compares the data in each section with the threshold, and finds out the points that exceed the threshold. These points are called threshold reference points. Based on the distance between the threshold reference point and the initial point of the section, calculate how much time is needed to adjust the error, that is, the error adjustment duration. On the basis of this duration, determine a control time point before the initial point of the section to ensure that the control strategy can be implemented before the error occurs. For example, an error adjustment duration can be determined forward from the initial point of the section to obtain the control time point. For example, if the distance between the threshold reference point and the initial point of the section is small, advance control is performed at a time point with a larger distance before the initial point of the section to prevent the situation where the control is not completed in time when the threshold reference point is reached.

[0061] In one embodiment of the present application, based on the error value data within the segment, an error value fluctuation sequence corresponding to each segment is generated. The regional extreme point is determined in the error value fluctuation sequence, and the minimum value in the sequence after the regional extreme point is determined, and an error detection sequence is constructed based on the minimum value. The regional extreme point is detected based on the error detection sequence, and a reference sequence is constructed based on the regional extreme point that meets the detection conditions. The characteristic points of the reference sequence are determined, and the sequence floating pattern of the reference sequence is determined based on the characteristic points. Based on the sequence floating patterns corresponding to multiple segments, a sequence floating pattern set is constructed, and the sequence floating pattern set is matched with a preset error type library to determine the error type; wherein the preset error type library includes multiple floating pattern sets, and also includes error types corresponding to multiple floating pattern sets.

[0062] Specifically, the embodiment of the present application determines the specific process of the error type based on the error values ​​corresponding to multiple segments respectively: for each segment, an error value fluctuation sequence is generated based on the error value data in the segment. This sequence describes the change of the error value over time. Assuming that the error value data composed of multiple segments is [0.1, 0.2, 0.15, 0.3, 0.25, 0.2, 0.18], then the corresponding error value fluctuation sequence is these values ​​themselves. In the error value fluctuation sequence, all local maximum points and local minimum points are determined, and then the minimum value is found in the sequence after each regional extreme point. For example, in the above error value fluctuation sequence, assuming that the local maximum point is 0.3, then in the sequence [0.25, 0.2, 0.18] after it, the minimum value is 0.18.

[0063] Furthermore, based on the minimum value after each regional extreme point, an error detection sequence is constructed. This sequence is used for the subsequent detection of regional extreme points. The regional extreme points are detected using the error detection sequence to screen out regional extreme points that meet specific conditions (such as fluctuation amplitude, duration, etc.), and these points will constitute a reference sequence. In the reference sequence, feature points are found, which represent the main change characteristics of the sequence. Then, the sequence floating pattern of the reference sequence is determined based on these feature points. For example, feature points are inflection points, extreme points, etc. in the sequence. The sequence floating pattern describes the relative position and change trend between these feature points. The determined sequence floating pattern is matched with the preset error type library to determine the error type, wherein the preset error type library in the embodiment of the present application includes multiple error types, and also includes sequence floating patterns corresponding to multiple error types.

[0064] Step 106: When the structural data of the air outlet meets the first preset structural data, a spinning grooving instruction and a second starting instruction of the spinning machine are received to produce the remaining part of the liquid reservoir.

[0065] In one embodiment of the present application, when the stamping and shaping of the outlet end is completed and the structural data of the outlet end after processing meets the requirements, a welding instruction will be received. At this time, the outlet elbow is installed in the outlet elbow installation hole, and the welding flange on the outlet elbow body is used to assist in positioning, ensuring that the straight pipe part of the outlet elbow and the cylinder body are coaxially arranged with the outlet elbow installation hole and the cylinder body, and the outlet elbow and the outlet elbow installation hole are welded and fixed by the resistance welding process to complete the installation of the outlet elbow.

[0066] In one embodiment of the present application, the middle partition is installed to a first preset position along the axial direction of the cylinder blank. A first spinning machine grooving instruction is received, and the grooving machine is used to groove the outer wall of the cylinder blank corresponding to the first preset position, so that the outer wall of the cylinder blank protrudes radially toward the inside of the cylinder along the cylinder, and the middle partition is axially limited. The filter screen is installed to a second preset position along the axial direction of the cylinder blank. A second spinning machine grooving instruction is received, and the grooving machine is used to groove the outer wall of the cylinder blank corresponding to the second preset position, so that the outer wall of the cylinder blank protrudes radially toward the inside of the cylinder along the cylinder, and the filter screen is axially limited.

[0067] Specifically, the middle partition is first installed at the first preset position inside the cylinder, and the first spinning machine groove instruction is received. The groove cutting machine grooves the outer wall of the cylinder so that the outer wall of the cylinder protrudes toward the inside of the cylinder along the radial direction of the cylinder, and the middle partition is axially limited. Then, the filter is installed at the second preset position inside the cylinder, and the second spinning machine groove instruction is received. The groove cutting machine also grooves the outer wall of the cylinder so that the outer wall of the cylinder protrudes toward the inside of the cylinder along the radial direction of the cylinder, and the filter is axially limited.

[0068] In one embodiment of the present application, after the middle partition and the filter are installed, the second start-up instruction of the spinning machine is received, and the cylinder is reversibly installed on the pressure spindle of the spinning machine. The pressure spindle drives the cylinder to rotate and cooperates with the spinning wheel to spin the other end of the cylinder to form an air intake end. When the structural data of the air intake end meets the second preset structural data, the air intake straight pipe is welded to the air intake end to realize the digital production of the air conditioner compressor accumulator.

[0069] In one embodiment of the present application, when the structural data of the gas outlet end meets the first preset structural data, the first instruction for cleaning the inside of the barrel blank is started. After the gas outlet elbow is welded to the gas outlet end, the second instruction for cleaning the inside of the barrel blank is started. After spinning and cutting grooves at a preset position for axial limiting, the third instruction for cleaning the inside of the barrel blank is started. When the structural data of the gas inlet end meets the second preset structural data, the fourth instruction for cleaning the inside of the barrel blank is started.

[0070] Specifically, if the structural data of the gas outlet meets the first preset structural data, the first instruction for cleaning the inside of the barrel blank is started to ensure that the inside of the barrel blank reaches a certain cleaning standard before welding or other processing steps. After welding is completed, the second instruction for cleaning the inside of the barrel blank is started to remove impurities or contaminants that may be generated during the welding process to ensure the smooth progress of subsequent processing steps. After the spinning grooving is completed, the third instruction for cleaning the inside of the barrel blank is started to remove metal chips or other tiny particles generated during the spinning grooving process to ensure the cleanliness of the inside of the barrel blank. If the structural data of the air inlet end meets the second preset structural data, the fourth instruction for cleaning the inside of the barrel blank is started to ensure that the barrel blank reaches the required cleaning standard before final assembly or delivery.

[0071] As a feasible implementation method, Figure 2 A flow chart of a liquid reservoir processing provided in an embodiment of the present application is as follows: Figure 2 As shown in the first and second figures, the pressure spindle drives the barrel 1 blank to move and contact with the bottom die, drives the upper die to move, and cooperates with the bottom die to shape the initial outlet end so that the bottom of the initial outlet end is flat, and punches out the outlet bend pipe installation hole 111 for installing the outlet bend pipe 21 in the center of the initial outlet end, completing the processing of the outlet end 11, and the outlet end 11 includes a flat portion 112 and an arc portion 113. The outlet bend pipe 21 is installed in the outlet bend pipe installation hole 111, and the welding flange 23 on the pipe body of the outlet bend pipe 21 assists in positioning, ensuring that the straight pipe part of the outlet bend pipe 21 that cooperates with the barrel 1 is coaxially arranged with the outlet bend pipe installation hole 111 and the barrel 1, and the outlet bend pipe 21 is welded and fixed to the outlet bend pipe installation hole 111 by a resistance welding process, completing the installation of the outlet bend pipe 21.

[0072] like Figure 2 As shown in the third figure, the middle partition 4 is installed inside the cylinder 1 along the axial direction of the cylinder 1, and a groove is cut on the outer wall of the cylinder 1 by a groove cutter so that the outer wall of the cylinder 1 protrudes toward the inside of the cylinder 1 along the radial direction of the cylinder 1, and the middle partition 4 is axially limited. Then, the filter screen 3 is installed inside the cylinder 1 along the axial direction of the cylinder 1, and a groove is cut on the outer wall of the cylinder 1 by a groove cutter so that the outer wall of the cylinder 1 protrudes toward the inside of the cylinder 1 along the radial direction of the cylinder 1, and the filter screen 3 is axially limited.

[0073] like Figure 2As shown in the fourth and fifth figures, the pressure spindle drives the cylinder 1 to rotate and cooperates with the spinning wheel to spin the other end of the cylinder 1 to form the air inlet end 12. The air inlet straight pipe 22 is installed in the air inlet straight pipe mounting hole 121, and the welding flange 23 on the air inlet straight pipe 22 is used to assist in positioning to ensure that the air inlet straight pipe 22 is coaxial with the air inlet straight pipe mounting hole 121 and the cylinder 1, and the air inlet straight pipe 22 is welded and fixed to the air inlet straight pipe mounting hole 121 by resistance welding to complete the installation of the air inlet straight pipe 22. Then the entire digital production and processing of the liquid reservoir is completed. The welding flange 23 of the air inlet straight pipe 22 is provided with a first limiting step 231 adapted to the air inlet straight pipe mounting hole 121 on one side facing the air inlet straight pipe mounting hole 121, and the welding flange 23 of the air outlet elbow 21 is provided with a second limiting step 232 adapted to the air outlet elbow mounting hole 111 on both sides facing the air outlet elbow mounting hole 111. The arrangement of the first limiting step 231 and the second limiting step 232 facilitates the positioning and installation of the air inlet straight pipe 22 and the air outlet elbow 21 , and also facilitates fixation during welding, thereby ensuring welding effects and improving processing and production efficiency.

[0074] Step 107, obtaining production information corresponding to different production links of the liquid reservoir, and mapping the production information corresponding to different production links with the product code, so as to trace the production process of the liquid reservoir through the mapping relationship.

[0075] In one embodiment of the present application, production information is uploaded to the blockchain. Based on the data type of the production information, the production information is divided into general information and sensitive information, and the sensitive information is homomorphically encrypted to obtain encrypted sensitive information. The encrypted sensitive information and the general information are hashed to generate a hash value that uniquely corresponds to the production information of the current production link. Based on the production information and the hash value, the processing information codes corresponding to different production links of the liquid reservoir are generated. Based on the production sequence, a connection network is constructed between adjacent production links, and the processing information codes corresponding to different production links are associated based on the connection network to construct a relationship association network corresponding to the processing information code. The relationship association network is mapped to the product code to trace the liquid reservoir product through the mapping relationship.

[0076] Specifically, after uploading the production information of the liquid reservoir to the blockchain, the production information is divided into general information and sensitive information, where general information can be: information that does not involve core secrets such as production date, production batch, production line number, etc., and sensitive information can be: information that needs to be kept confidential such as production process parameters and raw material supplier information. Sensitive information is homomorphically encrypted to protect the security of sensitive information during transmission and storage, while allowing necessary calculations and analysis in an encrypted state. The encrypted sensitive information and general information are hashed, and a hash value that uniquely corresponds to the production information of the current production link can be generated through hashing. Based on the production information and hash value, the processing information codes corresponding to different production links of the liquid reservoir are generated. This code is unique and can accurately identify the production information of the liquid reservoir in different production links.

[0077] Furthermore, based on the production sequence, a connection network is constructed between adjacent production links. This network can clearly show the entire process of the liquid reservoir from raw materials to finished products, as well as the association between each production link. Based on the connection network, the processing information codes corresponding to different production links are associated to construct a complete relationship association network corresponding to the processing information codes, which contains all the key information of the liquid reservoir in the production process. The relationship association network is mapped with the product code. The product code is the unique identifier of the liquid reservoir, through which the corresponding production information can be quickly found in the relationship association network. When it is necessary to trace the production information of a certain liquid reservoir, only the product code is needed, and all the processing information codes and corresponding hash values ​​of the liquid reservoir in different production links can be queried on the blockchain through the mapping relationship, thereby realizing product traceability.

[0078] Figure 3 This is a schematic diagram of the structure of a digital production system for an air-conditioning compressor accumulator provided in an embodiment of the present application. Figure 3As shown, the digital production system 200 of the air-conditioning compressor liquid reservoir includes: a liquid reservoir preparation unit 201, a spinning unit 202, a spinning data acquisition and processing unit 203, a remaining production unit 204 and a traceability unit 205; the liquid reservoir preparation unit 201 is used to obtain the liquid reservoir preparation completion information and generate a product code; wherein the liquid reservoir preparation includes a barrel blank, an air inlet straight pipe, an air outlet elbow, a middle partition and a filter; the spinning unit 202 is used to receive the first start-up instruction of the spinning machine to start the spinning machine to perform spinning and stamping shaping on the air outlet end of the barrel blank; the spinning data acquisition and processing unit 203 is used to receive the spinning data acquisition and processing instruction, divide the collected radian angle data of the spinning closing of the liquid reservoir end into intervals, and divide the radian angle data into intervals based on the radian angle data differences corresponding to different intervals. The actual fitting curve is fitted based on the data, and the actual fitting curve is adjusted based on the tube and shell data of the liquid reservoir; the spinning data acquisition and processing unit 203 is also used to obtain the operating data of the spinning machine, generate the spinning machine operating fitting curves corresponding to different operating conditions based on the spinning machine operating data, and pre-regulate the spinning machine based on the adjusted actual fitting curve and the spinning machine operating fitting curve; the remaining part production unit 204 is used to receive the spinning grooving instruction and the second starting instruction of the spinning machine when the structural data of the air outlet end meets the first preset structural data, and produce the remaining part of the liquid reservoir; the tracing unit 205 is used to obtain the production information corresponding to different production links of the liquid reservoir, and map the production information corresponding to different production links with the product code, so as to trace the production process of the liquid reservoir through the mapping relationship.

[0079] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0080] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the embodiments of the present application may have various changes and variations. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A digital production method for an air-conditioning compressor accumulator, characterized in that: The method comprises: Obtaining the completion information of the liquid reservoir material preparation and generating a product code; wherein the liquid reservoir material preparation includes a barrel blank, an air inlet straight pipe, an air outlet elbow pipe, a middle partition plate and a filter screen; receiving a first start-up instruction of a spinning machine, and performing a spinning process and a stamping process on the outlet end of the barrel blank through the spinning machine; Receive a spinning data collection and processing instruction, divide the collected radian angle data of the spinning closing of the end of the liquid reservoir into intervals, fit the radian angle data based on the radian angle data differences corresponding to different intervals, obtain an actual fitting curve, and adjust the actual fitting curve based on the tube shell data of the liquid reservoir; Acquire the spinning machine operation data, and generate the spinning machine operation fitting curves corresponding to different operation conditions respectively based on the spinning machine operation data; Pre-regulating the spinning machine based on the adjusted actual fitting curve and the spinning machine operation fitting curve; When the structural data of the gas outlet end meets the first preset structural data, a spinning grooving instruction and a second starting instruction of the spinning machine are received to produce the remaining part of the liquid reservoir; The production information corresponding to different production links of the liquid reservoir is obtained, and the production information corresponding to the different production links is mapped with the product code, so as to trace the production process of the liquid reservoir through the mapping relationship.

2. The digital production method of an air-conditioning compressor accumulator according to claim 1, characterized in that: The production information corresponding to the different production links is mapped with the product code to trace the production process of the liquid reservoir through the mapping relationship, specifically including: Uploading the production information to the blockchain; Based on the data type of the production information, the production information is divided into common information and sensitive information, and the sensitive information is homomorphically encrypted to obtain encrypted sensitive information; Performing hash calculation on the encrypted sensitive information and the general information to generate a hash value uniquely corresponding to the production information of the current production link; Based on the production information and the hash value, generate processing information codes corresponding to different production links of the liquid reservoir; Based on the production sequence, a connection network is constructed between adjacent production links, and based on the connection network, the processing information codes corresponding to different production links are associated to construct a relationship association network corresponding to the processing information codes; The relationship association network is mapped with the product code to trace the source of the liquid reservoir product through the mapping relationship.

3. The digital production method of an air-conditioning compressor accumulator according to claim 1, characterized in that: The collected radian angle data of the spinning and closing of the end of the liquid reservoir are divided into intervals, and the radian angle data are fitted based on the radian angle data differences corresponding to different intervals to obtain an actual fitting curve, specifically including: The radian angle data of the spinning closing is obtained by using angle sensors arranged at different positions of the end of the liquid reservoir; According to the shape of the spinning closing end of the liquid reservoir, the arc angle data of the spinning closing end is divided into a plurality of first intervals; Acquire adjacent change differences of the radian angle data in the first interval, determine interval connection points based on the adjacent change differences, and divide the first interval into a plurality of second intervals based on the interval connection points; Determine interval radian differences corresponding to a plurality of the second intervals respectively, and perform polynomial matching of different orders on the plurality of the second intervals based on the interval radian differences; Based on polynomials of different orders corresponding to a plurality of the second intervals, and the radian angle data of the spinning closure corresponding to a plurality of the second intervals, fitting is performed to obtain the actual fitting curve.

4. The digital production method of an air-conditioning compressor accumulator according to claim 1, characterized in that: The adjusting of the actual fitting curve based on the tube and shell data of the liquid reservoir specifically includes: Acquire a tube shell thickness in the tube shell data, and determine a first adjustment coefficient based on a difference between the tube shell thickness and a reference tube shell thickness; Based on the relationship between the historical shell and tube clamping vertical angles in the historical shell and tube data and the adjustment coefficients of the historical fitting curve, a clamping error detection model is constructed; Acquire a tube shell clamping vertical angle in the tube shell data, input the tube shell clamping vertical angle into the clamping error detection model, and obtain a second adjustment coefficient; The actual fitting curve is adjusted based on the first adjustment coefficient and the second adjustment coefficient.

5. The digital production method of an air-conditioning compressor accumulator according to claim 1, characterized in that: The step of acquiring the spinning machine operation data and generating spinning machine operation fitting curves corresponding to different operation conditions based on the spinning machine operation data specifically includes: Acquire operating data of the spinning machine; wherein the operating data at least includes temperature, pressure and rotation speed; Fitting the operating data by statistical methods to obtain fitting curve groups corresponding to different operating conditions; Normalizing the operating data of the spinning machine and the abscissa of the fitting curve group; Determine a reference curve in the fitting curve group, compare other curves in the fitting curve group with the reference curve, and determine, in the other curves, relative values ​​of ordinates corresponding to the same reference points as the reference curve; Based on the relative value of the ordinate and the normalized data, a fitting curve of the spinning machine operation corresponding to the spinning machine under different operating conditions is obtained.

6. The digital production method of an air-conditioning compressor accumulator according to claim 1, characterized in that: The step of controlling the spinning machine in advance based on the adjusted actual fitting curve and the spinning machine operation fitting curve specifically includes: Extracting key parameters of the adjusted actual fitting curve; wherein the extracted features include at least one of the slope of the curve, the curvature of the curve, and the extreme point of the curve; Inputting the extracted key parameters and the corresponding spinning machine operation fitting curves under different operating conditions into a spinning prediction model, so as to output a prediction fitting curve corresponding to a future time period through the spinning prediction model; Determine the first contribution value of different operation data features to the prediction fitting curve corresponding to the future time period by the SHAP algorithm, and draw a feature global explanation map based on the contribution value; A local linear model is constructed by using the LIME algorithm, second contribution values ​​of different operation data features to the prediction fitting curve corresponding to the future time period are determined by using the weights of the local linear model, and a feature local explanation map is drawn based on the second contribution values; Based on the feature global interpretation map and the feature local interpretation map, a future spinning prediction fitting curve is constructed, and the future spinning prediction fitting curve is compared with a target curve, and the spinning machine is adjusted in advance based on the comparison result.

7. The digital production method of the air conditioner compressor accumulator according to claim 6, characterized in that: The comparing the future spinning prediction fitting curve with the target curve and adjusting the spinning machine in advance based on the comparison result specifically includes: Comparing the future spinning prediction fitting curve with the target curve to determine an error point between the two, and dividing the future spinning prediction fitting curve into a plurality of sections based on the error point; Based on the error values ​​respectively corresponding to the plurality of sections, the error type is determined; wherein the error type includes position error, shape error and size error; Inputting the error type and the error value into a preset error control model, and outputting a corresponding spinning machine control strategy through the preset error control model; Obtaining a spinning error threshold value for the end of the reservoir, comparing the data in the section with the spinning error threshold value, and determining a threshold reference point; Based on the distance between the threshold reference point and the initial point of the section, the error adjustment duration is determined, and based on the adjustment duration, a control time point is determined before the initial point of the section, so as to implement the spinning machine control strategy at the control time point.

8. The digital production method of the air conditioner compressor accumulator according to claim 7, characterized in that: The determining of the error type based on the error values ​​respectively corresponding to the plurality of segments specifically includes: Based on the error value data in the segment, the error value fluctuation sequence corresponding to each segment is generated; Determine a regional extreme point in the error value fluctuation sequence, determine a minimum value in the sequence after the regional extreme point, and construct an error detection sequence based on the minimum value; Detecting the regional extreme value points based on the error detection sequence, and constructing a reference sequence based on the regional extreme value points that meet the detection conditions; Determining characteristic points of the reference sequence, and determining a sequence floating mode of the reference sequence based on the characteristic points; Based on the sequence floating patterns corresponding to the multiple segments respectively, a sequence floating pattern set is constructed, and the sequence floating pattern set is matched with a preset error type library to determine the error type; wherein the preset error type library includes multiple floating pattern sets, and also includes multiple error types corresponding to the floating pattern sets respectively.

9. The digital production method of an air-conditioning compressor accumulator according to claim 1, characterized in that: When the structural data of the gas outlet end meets the first preset structural data, a spinning groove engraving instruction and a second starting instruction of the spinning machine are received to produce the remaining part of the liquid reservoir, specifically including: When the structural data of the gas outlet end meets the first preset structural data, welding the gas outlet elbow to the gas outlet end; Installing the middle partition plate to a first preset position along the axial direction of the barrel blank; Receiving a notching instruction from a first spinning machine, notching the outer wall of the barrel blank corresponding to the first preset position by the notching machine, so that the outer wall of the barrel blank protrudes radially toward the inside of the barrel, and axially limits the middle partition; Installing the filter screen to a second preset position along the axial direction of the barrel blank; Receiving a notching instruction from a second spinning machine, notching the outer wall of the barrel blank corresponding to the second preset position by the notching machine, so that the outer wall of the barrel blank protrudes radially toward the inside of the barrel, and axially limiting the filter screen; Receive the second start-up instruction of the spinning machine, perform spinning processing on the air inlet end of the cylinder blank through the spinning machine, and when the structural data of the air inlet end meets the second preset structural data, weld the air inlet straight pipe to the air inlet end to realize the digital production of the air-conditioning compressor liquid reservoir.

10. A digital production system for an air-conditioning compressor accumulator, characterized in that: The system includes a liquid reservoir preparation unit, a spinning unit, a spinning data acquisition and processing unit, a remaining part production unit, and a traceability unit; The liquid reservoir material preparation unit is used to obtain the liquid reservoir material preparation completion information and generate a product code; wherein the liquid reservoir material preparation includes a barrel blank, an air inlet straight pipe, an air outlet elbow, a middle partition plate and a filter screen; The spinning unit is used to receive a first start instruction of the spinning machine to start the spinning machine to perform spinning and stamping shaping on the outlet end of the barrel blank; The spinning data acquisition and processing unit is used to receive a spinning data acquisition and processing instruction, divide the collected radian angle data of the spinning closing of the end of the liquid reservoir into intervals, fit the radian angle data based on the radian angle data differences corresponding to different intervals, obtain an actual fitting curve, and adjust the actual fitting curve based on the tube shell data of the liquid reservoir; The spinning data acquisition and processing unit is further used to obtain the spinning machine operation data, generate the spinning machine operation fitting curves corresponding to different operation conditions based on the spinning machine operation data, and perform early control on the spinning machine based on the adjusted actual fitting curve and the spinning machine operation fitting curve; The remaining part production unit is used to receive the spinning grooving instruction and the second starting instruction of the spinning machine when the structural data of the gas outlet end meets the first preset structural data, and produce the remaining part of the liquid reservoir; The tracing unit is used to obtain the production information corresponding to different production links of the liquid reservoir, and map the production information corresponding to the different production links with the product code, so as to trace the production process of the liquid reservoir through the mapping relationship.

Images