Press machine fault early warning method

Abstract

The invention relates to a press machine fault early warning method, which comprises the following steps of: firstly collecting press machine processing data of different types of parts to obtain a plurality of sample data, further performing pressure value-time curve construction on all pressure data in each sample data, and then performing pressure value-time curve construction on all pressure data in each sample data; all the pressure value-time curves in each piece of sample data are fused into a coordinate system, and a fusion curve corresponding to each piece of sample data is obtained; cutting the fusion curve based on different working states of the press machine, calculating the change rate of the pressure corresponding to the press machine in different working states through the cut curve, and constructing and training a press machine wear model according to the change rate; and the prediction time of the wear fault of the press can be speculated by acquiring the real-time pressure data of the pressure.

Description

Technical Field

[0001] The present application relates to the technical field of presses, and in particular to a press failure early warning method. Background Art

[0002] A press (including punch presses and hydraulic presses) is a versatile, compact machine. Its versatility and high production efficiency make it suitable for a wide range of applications, including cutting, punching, blanking, bending, riveting, and forming. Presses apply strong pressure to metal blanks, causing them to plastically deform and fracture, ultimately creating parts.

[0003] Currently, presses often repeatedly process different types of workpieces during operation, using varying pressures for different parts. Consequently, the pressures provided by the press's power system also vary, leading to varying degrees of wear on the power system. Traditional technologies are unable to provide early warning of wear on the power systems of presses with complex operating ranges, often relying solely on operator experience, resulting in inaccurate predictions. Summary of the Invention

[0004] Based on this, it is necessary to provide a press failure warning method to address the problem that traditional presses with complex working ranges cannot provide early warning for wear and tear, and can only rely on the experience of staff, and their manual prediction results are not accurate.

[0005] The present application provides a press failure early warning method, comprising: Acquire a plurality of sample data, wherein the sample data includes a plurality of pressure data and a working state of a press; Constructing a pressure value-time curve for each pressure data based on each pressure data in each sample data; Construct a fusion curve based on all the pressure value-time curves in each sample data; The fusion curve is trimmed based on the working state of the press in each sample data to obtain multiple characteristic curves; Calculate the pressure change rate of the press under different working conditions based on each characteristic curve in each sample data; Construct and train a press wear model based on each sample data, each press pressure data, each pressure value-time curve, each fusion curve, each characteristic curve, and the time-based change rate of the press pressure under different working conditions as training data; Get a sample data to be tested; Parsing the sample data to be tested to obtain a plurality of pressure data to be tested in the sample data to be tested; A plurality of pressure data to be measured is input into a press wear model, the press wear model is started, and a predicted time of press wear failure output by the press wear model is obtained.

[0006] Furthermore, the pressure data includes: press working pressure data.

[0007] Furthermore, constructing a pressure value-time curve of each pressure data based on each pressure data in each sample data includes: Select a pressure data in a sample data; Obtaining the press working pressure data from the pressure data; Create a press working pressure value-time coordinate system; Filling the press working pressure data in the pressure data into the press working pressure value-time coordinate system to obtain a pressure value-time curve; Return to the process of selecting each pressure data in a sample data until each pressure data in each sample data has been selected once.

[0008] Furthermore, the step of constructing a fusion curve based on all pressure value-time curves in each sample data includes: Select a sample data; Obtain all pressure value-time curves in the sample data; Based on the time dimension, all pressure value-time curves are fused into a coordinate system to obtain a fused curve; Return to the process of selecting a sample data until each sample data has been selected once.

[0009] Furthermore, the fusion curve is clipped based on the working state of the press in each sample data to obtain multiple characteristic curves, including: Select a sample data; Acquire a working state of the press in the sample data, where the working state of the press includes approaching a workpiece, squeezing the workpiece, and resetting; Obtain the time range of approaching the workpiece, the time range of squeezing the workpiece, and the time range of resetting; Return to the process of selecting a sample data until each sample data has been selected once.

[0010] Furthermore, the method of clipping the fusion curve based on the working state of the press in each sample data to obtain multiple characteristic curves also includes: Select a sample data; Obtaining a time range of approaching the workpiece, a time range of squeezing the workpiece, and a time range of resetting in the sample data; The fusion curve is clipped based on a time range close to the workpiece to obtain a first characteristic curve; The fusion curve is trimmed based on the time range of the extruded workpiece to obtain a second characteristic curve; The fusion curve is trimmed based on the reset time range to obtain a third characteristic curve; Return to the process of selecting a sample data until each sample data has been selected once.

[0011] Furthermore, the calculation of the pressure change rate of the press under different working states based on each characteristic curve in each sample data includes: Selecting a first characteristic curve from a sample data; Obtaining all first segmented curves of the first characteristic curve; Select a first segment curve; Calculating the difference between the two endpoints of the first segmented curve, and defining the obtained difference as a first change value; Returning to the step of selecting a first segmented curve until each first segmented curve has been selected once, thereby obtaining a plurality of first change values; Calculate the average value of all first change values, and the obtained average value is defined as the pressure change rate of the press when it is close to the workpiece; Return to the step of selecting a first characteristic curve from a sample data until the first characteristic curve from each sample data has been selected once.

[0012] Furthermore, the calculation of the pressure change rate of the press under different working states based on each characteristic curve in each sample data further includes: Select a second characteristic curve from the sample data; Obtaining all second segmented curves of the second characteristic curve; Select a second segmented curve; Calculating the difference between the two endpoints of the second segmented curve, and defining the obtained difference as a second change value; Returning to the step of selecting a second segmented curve until each second segmented curve has been selected once, thereby obtaining a plurality of second change values; Calculate the average value of all second change values, and the obtained average value is defined as the pressure change rate of the press when the press is in a state of squeezing the workpiece; Return to the step of selecting a second characteristic curve from a sample data until the second characteristic curve from each sample data has been selected once.

[0013] Furthermore, the calculation of the pressure change rate of the press under different working states based on each characteristic curve in each sample data further includes: Select the third characteristic curve in a sample data; Obtaining all third segmented curves of the third characteristic curve; Select a third segment curve; Calculating the difference between the two endpoints of the third segmented curve, and defining the obtained difference as a third change value; Returning to the step of selecting a third segmented curve until each third segmented curve has been selected once, thereby obtaining a plurality of third change values; Calculate the average value of all third change values, and the obtained average value is defined as the pressure change rate of the press in the reset state; Return to the step of selecting the third characteristic curve in a sample data until the third characteristic curve in each sample data has been selected once.

[0014] Furthermore, the step of inputting the pressure data to be measured into the press wear model, starting the press wear model, and obtaining the predicted time of the press wear failure output by the press wear model includes: Analyze each pressure data to be measured to obtain a pressure value-time curve in each pressure data to be measured; Input each pressure value-time curve to be measured into the press wear model in chronological order, and obtain a first change value, a second change value, and a third change value output by the press wear model; constructing a pressure value-time prediction curve based on the obtained first change value, second change value, and third change value; Get the preset pressure value; The intersection of the preset pressure value and the pressure value-time prediction curve is calculated, and the time coordinate of the intersection is the wear failure prediction time.

[0015] The present application relates to a press failure early warning method, which first collects press processing data of different types of parts to obtain multiple sample data, and then constructs a pressure value-time curve for all pressure data in each sample data. Then, all pressure value-time curves in each sample data are fused into a coordinate system to obtain a fused curve corresponding to each sample data. The fused curve is then clipped based on the different working states of the press, and the rate of change of the pressure corresponding to the press under different working states is calculated using the clipped curve. A press wear model is then constructed and trained, and the predicted time of press wear failure can be inferred by obtaining real-time pressure data. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flow chart of a press machine failure warning method provided in one embodiment of the present application. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solutions and advantages of this application more clearly understood, the present application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0018] like Figure 1 As shown, in one embodiment of the present application, the press machine failure early warning method includes the following steps S100 to S900: S100 , obtaining a plurality of sample data, wherein the sample data includes a plurality of pressure data and a working state of a press.

[0019] Specifically, the plurality of sample data refers to data of processing different types of workpieces by the same press, and the same sample data refers to data of processing multiple workpieces of the same type.

[0020] The pressure data includes: press working pressure data, and the pressure data is hydraulic pressure data or pneumatic pressure data of the power system of the press; further, the wear of the power system of the press is often directly reflected in the pressure value change of the power system.

[0021] S200 , constructing a pressure value-time curve of each pressure data based on each pressure data in each sample data.

[0022] S300: Construct a fusion curve based on all pressure value-time curves in each sample data.

[0023] S400 , cutting the fusion curve based on the working state of the press in each sample data to obtain multiple characteristic curves.

[0024] Specifically, the working states of the press include approaching the workpiece, squeezing the workpiece, and resetting.

[0025] S500 , calculating the pressure change rate of the press under different working states based on each characteristic curve in each sample data.

[0026] S600, based on each sample data, each press pressure data, each pressure value-time curve, each fusion curve, each characteristic curve and the time-based change rate of the press pressure under different working conditions as training data, build and train the press wear model.

[0027] Specifically, a press wear model is constructed.

[0028] The press wear model is trained using each sample data point, each press pressure data point, each pressure-time curve, each fusion curve, each characteristic curve, and the time-dependent rate of change of the press pressure under different operating conditions as training data. This yields a trained press wear model. The trained press wear model can predict press wear based on the press pressure data.

[0029] S700: Obtain sample data to be tested.

[0030] S800: parsing the sample data to be tested to obtain a plurality of pressure data to be tested in the sample data to be tested.

[0031] S900: Input a plurality of pressure data to be measured into a press wear model, start the press wear model, and obtain a predicted time of press wear failure output by the press wear model.

[0032] In this embodiment, press processing data for different types of parts is first collected to obtain multiple sample data. Then, a pressure-time curve is constructed for all pressure data in each sample data. All pressure-time curves in each sample data are then fused into a coordinate system to obtain a fused curve corresponding to each sample data. The fused curves are then clipped based on the press's different operating states. The clipped curves are then used to calculate the rate of change of pressure under different operating states. This is used to construct and train a press wear model. This allows the predicted time of press wear failure to be inferred by acquiring real-time pressure data.

[0033] In one embodiment of the present application, constructing a pressure value-time curve of each pressure data based on each pressure data in each sample data includes the following steps S201 to S205: S201, selecting a pressure data from a sample data.

[0034] S202, obtaining the press working pressure data in the pressure data.

[0035] S203: Create a press working pressure value-time coordinate system.

[0036] S204 , filling the press working pressure data in the pressure data into the press working pressure value-time coordinate system to obtain a pressure value-time curve.

[0037] S205 , returning to the process of selecting each pressure data in a sample data until each pressure data in each sample data has been selected once.

[0038] In this embodiment, a press working pressure value-time coordinate system is created based on each pressure data, and then each pressure data is filled into the corresponding press working pressure value-time coordinate system based on the time dimension, and multiple points are obtained in the coordinate system. The multiple points are connected in sequence through a smooth curve to obtain a pressure value-time curve.

[0039] In one embodiment of the present application, the step of constructing a fusion curve based on all pressure value-time curves in each sample data includes the following steps S301 to S304: S301, select a sample data.

[0040] S302: Obtain all pressure value-time curves in the sample data.

[0041] S303: All pressure value-time curves are fused into a coordinate system based on the time dimension to obtain a fused curve.

[0042] S304, returning to the step of selecting a sample data, until each sample data has been selected once.

[0043] In this embodiment, each pressure data is recorded as a unit data, so the starting time node of each pressure data is 0. Based on this, the multiple pressure value-time curves obtained are fused into a coordinate system based on the time dimension to obtain a fused curve.

[0044] In one embodiment of the present application, the fusion curve is clipped based on the working state of the press in each sample data to obtain multiple characteristic curves, including the following S401 to S404: S401, select a sample data.

[0045] S402 , obtaining the working state of the press in the sample data, where the working state of the press includes approaching the workpiece, squeezing the workpiece, and resetting.

[0046] S403, obtaining a time range for approaching the workpiece, a time range for squeezing the workpiece, and a time range for resetting.

[0047] S404, returning to the step of selecting a sample data, until each sample data has been selected once.

[0048] The method of trimming the fusion curve based on the working state of the press in each sample data to obtain multiple characteristic curves also includes the following S405 to S410: S405: Select a sample data.

[0049] S406, obtaining a time range of approaching the workpiece, a time range of squeezing the workpiece, and a time range of resetting in the sample data.

[0050] S407 , cutting the fusion curve based on a time range close to the workpiece to obtain a first characteristic curve.

[0051] S408 , cutting the fusion curve based on the time range of extruding the workpiece to obtain a second characteristic curve.

[0052] S409: Clip the fusion curve based on the reset time range to obtain a third characteristic curve.

[0053] S410, returning to the step of selecting a sample data, until each sample data is selected once.

[0054] Specifically, in the present application, the time range for approaching the workpiece and the time range for resetting are both regarded as a fixed period of time; and the time range for squeezing the workpiece is determined based on the workpiece being processed.

[0055] In this embodiment, the same sample data represents the same workpiece being processed. Therefore, the press operating state corresponding to the same sample data is categorized into three stages. The fusion curve is then clipped based on the three stages to obtain characteristic curves corresponding to the press's different operating stages. This clipping step is based on the fact that the press experiences different degrees of wear under different operating states. By refining and distinguishing these differences, the accuracy of the final warning is improved.

[0056] In one embodiment of the present application, the calculation of the pressure change rate of the press under different working states based on each characteristic curve in each sample data includes the following steps S501 to S507: S501: Select a first characteristic curve from sample data.

[0057] S502: Acquire all first segmented curves of the first characteristic curve.

[0058] S503: Select a first segmented curve.

[0059] S504: Calculate the difference between the two endpoints of the first segmented curve, and define the obtained difference as a first change value.

[0060] S505 , returning to the step of selecting a first segmented curve, until each first segmented curve is selected once, and obtaining a plurality of first change values.

[0061] S506: Calculate an average value of all first change values, and define the obtained average value as the pressure change rate of the press when the press is close to the workpiece.

[0062] S507 , returning to the step of selecting a first characteristic curve from a sample data until the first characteristic curve from each sample data has been selected once.

[0063] The method of calculating the pressure change rate of the press under different working conditions based on each characteristic curve in each sample data further includes the following steps S100 to S514: S508: Select a second characteristic curve from the sample data.

[0064] S509: Acquire all second segmented curves of the second characteristic curve.

[0065] S510: Select a second segmented curve.

[0066] S511 , calculating the difference between the two endpoints of the second segmented curve, and defining the obtained difference as a second change value.

[0067] S512 , returning to the step of selecting a second segmented curve, until each second segmented curve is selected once, and obtaining a plurality of second change values.

[0068] S513, calculating an average value of all second change values, and the obtained average value is defined as the pressure change rate of the press when the press is in a state of extruding the workpiece.

[0069] S514 , returning to the step of selecting a second characteristic curve from a sample data until the second characteristic curve from each sample data has been selected once.

[0070] The method of calculating the pressure change rate of the press under different working conditions based on each characteristic curve in each sample data further includes the following steps S515 to S521: S515: Select a third characteristic curve from the sample data.

[0071] S516: Acquire all third segmented curves of the third characteristic curve.

[0072] S517: Select a third segmented curve.

[0073] S518: Calculate the difference between the two endpoints of the third segmented curve, and define the obtained difference as a third change value.

[0074] S519 , returning to the step of selecting a third segmented curve, until each third segmented curve is selected once, and obtaining a plurality of third change values.

[0075] S520: Calculate an average value of all third change values, and define the obtained average value as the pressure change rate of the press when it is in the reset state.

[0076] S521 , returning to the step of selecting a third characteristic curve from a sample data until the third characteristic curve from each sample data has been selected once.

[0077] In this embodiment, by analyzing and calculating the first characteristic curve, the second characteristic curve and the third characteristic curve in different sample data, the wear change of the press in different working states when processing different types of workpieces is obtained, so as to facilitate the wear prediction of subsequent wear of different types of workpieces.

[0078] In one embodiment of the present application, inputting the pressure data to be measured into the press wear model, starting the press wear model, and obtaining the predicted time of the press wear failure output by the press wear model includes the following steps S901 to S905: S901 , analyzing each piece of pressure data to be measured to obtain a pressure value-time curve in each piece of pressure data to be measured.

[0079] S902 , input each pressure value-time curve to be measured into the press wear model in chronological order, and obtain a first change value, a second change value, and a third change value output by the press wear model.

[0080] S903 : Construct a pressure value-time prediction curve based on the obtained first change value, second change value, and third change value.

[0081] S904: Obtain a preset pressure value.

[0082] S905 , calculating the intersection of the preset pressure value and the pressure value-time prediction curve, where the time coordinate of the intersection is the wear failure prediction time.

[0083] In this embodiment, the press wear model analyzes and calculates each piece of pressure data to obtain the first, second, and third change values for each sample. These values are then used to construct a pressure-time prediction curve for the workpiece type corresponding to the sample. The press's preset pressure value is then acquired, and the time coordinate information of the intersection of the preset pressure value and the pressure-time prediction curve is calculated. This intersection information represents the predicted wear failure time.

[0084] The various technical features of the above-described embodiments can be combined arbitrarily, and the execution order of the method steps is not restricted. In order to make the description concise, not all possible combinations of the various technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0085] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A press failure early warning method, characterized in that: The press failure early warning method comprises: Acquire a plurality of sample data, wherein the sample data includes a plurality of pressure data and a working state of a press; Constructing a pressure value-time curve for each pressure data based on each pressure data in each sample data; Construct a fusion curve based on all the pressure value-time curves in each sample data; The fusion curve is trimmed based on the working state of the press in each sample data to obtain multiple characteristic curves; Calculate the pressure change rate of the press under different working conditions based on each characteristic curve in each sample data; Construct and train a press wear model based on each sample data, each press pressure data, each pressure value-time curve, each fusion curve, each characteristic curve, and the time-based change rate of the press pressure under different working conditions as training data; Get a sample data to be tested; Parsing the sample data to be tested to obtain a plurality of pressure data to be tested in the sample data to be tested; A plurality of pressure data to be measured is input into a press wear model, the press wear model is started, and a predicted time of press wear failure output by the press wear model is obtained.

2. The press failure early warning method according to claim 1, characterized in that: The pressure data includes: press working pressure data.

3. The press failure early warning method according to claim 2, characterized in that: The step of constructing a pressure value-time curve of each pressure data based on each pressure data in each sample data includes: Select a pressure data in a sample data; Obtaining the press working pressure data from the pressure data; Create a press working pressure value-time coordinate system; Filling the press working pressure data in the pressure data into the press working pressure value-time coordinate system to obtain a press working pressure value-time curve; Return to the process of selecting each pressure data in a sample data until each pressure data in each sample data has been selected once.

4. The press failure early warning method according to claim 3, characterized in that: The method of constructing a fusion curve based on all pressure value-time curves in each sample data includes: Select a sample data; Obtain all press working pressure value-time curves in the sample data; Based on the time dimension, all the press working pressure value-time curves are fused into a coordinate system to obtain a fusion curve; Return to the process of selecting a sample data until each sample data has been selected once.

5. The press machine failure early warning method according to claim 4, characterized in that: The fusion curve is trimmed based on the working state of the press in each sample data to obtain multiple characteristic curves, including: Select a sample data; Acquire a working state of the press in the sample data, where the working state of the press includes approaching a workpiece, squeezing the workpiece, and resetting; Obtain the time range of approaching the workpiece, the time range of squeezing the workpiece, and the time range of resetting; Return to the process of selecting a sample data until each sample data has been selected once.

6. The press machine failure early warning method according to claim 5, characterized in that: The method further includes: cutting the fusion curve based on the working state of the press in each sample data to obtain multiple characteristic curves; Select a sample data; Obtaining a time range of approaching the workpiece, a time range of squeezing the workpiece, and a time range of resetting in the sample data; The fusion curve is clipped based on the time range close to the workpiece to obtain a first characteristic curve; The fusion curve is trimmed based on the time range of the extruded workpiece to obtain a second characteristic curve; The fusion curve is trimmed based on the reset time range to obtain a third characteristic curve; Return to the process of selecting a sample data until each sample data has been selected once.

7. The press failure early warning method according to claim 6, characterized in that: The calculation of the pressure change rate of the press under different working conditions based on each characteristic curve in each sample data includes: Selecting a first characteristic curve from a sample data; Obtaining all first segmented curves of the first characteristic curve; Select a first segment curve; Calculating the difference between the two endpoints of the first segmented curve, and defining the obtained difference as a first change value; Returning to the step of selecting a first segmented curve until each first segmented curve has been selected once, thereby obtaining a plurality of first change values; Calculate the average value of all first change values, and the obtained average value is defined as the pressure change rate of the press when it is close to the workpiece; Return to the step of selecting a first characteristic curve from a sample data until the first characteristic curve from each sample data has been selected once.

8. The press machine failure early warning method according to claim 7, characterized in that: The calculation of the pressure change rate of the press under different working conditions based on each characteristic curve in each sample data further includes: Select a second characteristic curve from the sample data; Obtaining all second segmented curves of the second characteristic curve; Select a second segmented curve; Calculating the difference between the two endpoints of the second segmented curve, and defining the obtained difference as a second change value; Returning to the step of selecting a second segmented curve until each second segmented curve has been selected once, thereby obtaining a plurality of second change values; Calculate the average value of all second change values, and the obtained average value is defined as the pressure change rate of the press when the press is in a state of squeezing the workpiece; Return to the step of selecting a second characteristic curve from a sample data until the second characteristic curve from each sample data has been selected once.

9. The press machine failure early warning method according to claim 8, characterized in that: The calculation of the pressure change rate of the press under different working conditions based on each characteristic curve in each sample data further includes: Select the third characteristic curve in a sample data; Obtaining all third segmented curves of the third characteristic curve; Select a third segment curve; Calculating the difference between the two endpoints of the third segmented curve, and defining the obtained difference as a third change value; Returning to the step of selecting a third segmented curve until each third segmented curve has been selected once, thereby obtaining a plurality of third change values; Calculate the average value of all third change values, and the obtained average value is defined as the pressure change rate of the press in the reset state; Return to the step of selecting the third characteristic curve in a sample data until the third characteristic curve in each sample data has been selected once.

10. The press machine failure early warning method according to claim 9, characterized in that: The step of inputting the pressure data to be measured into the press wear model, starting the press wear model, and obtaining the predicted time of the press wear failure output by the press wear model includes: Analyze each pressure data to be measured to obtain a pressure value-time curve in each pressure data to be measured; Input each pressure value-time curve to be measured into the press wear model in chronological order, and obtain a first change value, a second change value, and a third change value output by the press wear model; constructing a pressure value-time prediction curve based on the obtained first change value, second change value, and third change value; Get the preset pressure value; The intersection of the preset pressure value and the pressure value-time prediction curve is calculated, and the time coordinate of the intersection is the wear failure prediction time.

Images