Intelligent stamping die failure monitoring system and method

Abstract

The invention provides an intelligent stamping die failure monitoring system and method. The intelligent stamping die failure monitoring system comprises a data acquisition layer, a data processing layer and a data processing layer, wherein a voltage sensor is used for detecting voltage fluctuation of a stamping machine, a temperature sensor is used for detecting the temperature of cooling liquid of the stamping machine, a vibration sensor is used for detecting the machine table vibration frequency of the stamping machine, and a photoelectric encoder is used for acquiring stamping frequency data; aiming at the to-be-processed steel belt, arranging a magnetic flux detection device to detect the magnetic flux change of the steel belt; the data collected by the data acquisition layer is timestamped and then transmitted to the data storage layer, the mold failure is monitored by introducing magnetic flux parameters, a failure model is constructed, and the mold failure condition is monitored by setting a three-level early warning mechanism.

Description

Technical Field

[0001] This invention relates to the field of industrial control system technology, and in particular to an intelligent stamping die failure monitoring system and method. Background Technology

[0002] The traditional equipment manufacturing industry has begun to develop from the low-end and mid-range to the high-end, which puts forward higher requirements for the precision of parts. This largely depends on the precision and quality of molds. However, molds inevitably encounter various problems during repeated stamping processes, leading to mold failure. Therefore, mold failure is an important technical factor affecting the precision of parts, mold life and production costs.

[0003] The background of this patent is mainly the stamping of rolls of steel strip into individual perforated parts. During the stamping process, due to the uneven material of the steel strip and internal stress issues, some areas of the steel strip become abnormally hard, causing wear and even failure of the die during stamping, affecting product quality. In addition, excessively fast processing speed and insufficient cooling time of the stamping die can also lead to die failure. Voltage fluctuations in the factory cause fluctuations in the instantaneous stamping pressure of the die, which can also lead to die failure. Errors in the die manufacturing process by the die manufacturer and unreasonable material composition of the die can also cause die failure during use. Based on the above background, die failure involves multiple uncertain factors, making it difficult to control individually. Improving die lifespan and implementing monitoring are crucial. During the processing of parts, due to the high speed and continuous processing, random inspections of parts are used to indirectly monitor die failure. However, the inspection intervals are long, and by the time a quality problem is discovered, a batch of raw materials has already been wasted. Therefore, there is an urgent need for a monitoring system that can provide early warning to reduce raw material waste, and simultaneously monitor and adjust various factors to improve die monitoring efficiency. Summary of the Invention

[0004] Therefore, the present invention provides an intelligent stamping die failure monitoring system and method to achieve intelligent monitoring and alarm of die failure.

[0005] To achieve the above objectives, the present invention employs the following steps: a method for intelligent stamping die failure monitoring.

[0006] Step 1, data acquisition: Collect key data by setting up sensors;

[0007] Individual data collection is performed on each stamping machine, including voltage, coolant temperature, machine vibration frequency, and stamping frequency. A magnetic flux detection device is installed for the steel strip to be processed. Data is timestamped synchronously to ensure accurate analysis.

[0008] Step 2: Data storage, storing the acquired data;

[0009] The data mentioned above is stored locally and then transmitted to the backend server.

[0010] Step 3: Data is transmitted to the backend server for preprocessing;

[0011] Remove outliers from the raw data, standardize and normalize the data, and extract feature values;

[0012] Step 4: Construct a parameter-based failure prediction model, input parameters into the model, and calculate the current wear state:

[0013] Starting from a multi-parameter linear fusion model, and combining failure weight allocation and normalization factors, a comprehensive failure scoring model is formed; each term represents the normalized influence degree of a failure influencing factor, constructing a comprehensive failure factor F. fail Its expression is:

[0014]

[0015] Among them: the weight coefficients w1 to w5 are optimized monthly (based on regression analysis of historical failure cases).

[0016] w1 to w5 are weighting coefficients, reflecting the degree of influence of each parameter on failure, with a value range of [0,1]. The optimal values ​​are: w1 = 0.15, w2 = 0.25, w3 = 0.2, w4 = 0.25, and w5 = 0.15.

[0017] V0 is the nominal voltage, also known as the rated voltage of the equipment;

[0018] ΔV is the absolute value of the voltage fluctuation;

[0019] T cool The temperature of the coolant generally fluctuates between 20 and 60 degrees Celsius, with ≤40 being the optimal state.

[0020] T ref This is the reference temperature, representing a normal temperature baseline value; the set value is 20 degrees Celsius.

[0021] T max The maximum permissible temperature is set to a safety limit of 60 degrees Celsius.

[0022] f vib The vibration frequency is typically 0–1000 Hz; values ​​below 300 Hz are considered ideal.

[0023] f0 is the normal vibration frequency, which is the frequency within the allowable voltage fluctuation range of the equipment. This frequency is the next normal operating frequency for the equipment; in this system, it is set to 200 Hz.

[0024] f max This is the upper limit of the vibration frequency, and the threshold for the system to mark an anomaly; the value is 1000 Hz.

[0025] N press This refers to the actual number of stamping operations, typically ranging from 0 to 120 times per minute. Ideally, it should be below 100 times per minute.

[0026] N0 is the base frequency, a steady-state value of the device, which is 90 times per minute.

[0027] N max The maximum allowed frequency is set at 200 times per minute;

[0028] Φ0 is the reference magnetic flux; the normal magnetic flux value of the steel strip is 0.1~0.5wb.

[0029] ΔΦ is the difference between the actual detected magnetic flux and the reference magnetic flux;

[0030] Step 5: Set alarm thresholds and perform real-time monitoring, automatically recording historical data for model optimization;

[0031] The system is configured with a tiered early warning mechanism:

[0032] Level 1 warning (F) fail ∈[0.5,0.7)): The system displays "Mold Status Offset," prompting the operator to pay attention;

[0033] Level II warning (F) fail ∈[0.7,0.9)): The system prompts "Mold status abnormal", and it is recommended to check the cooling system and voltage stability;

[0034] Level III Warning (F) fail ≥0.9): The system will automatically stop, prompting you to replace the mold or repair the equipment;

[0035] Step Six: Set up the feedback and optimization module, and adjust the process parameters according to the model output;

[0036] Adjust the parameters in step one to improve the mold's lifespan;

[0037] Including F fail When the value is greater than 0.6, reduce the stamping speed, thereby reducing the stamping frequency;

[0038] When T cool When the temperature is above 45 degrees Celsius, increase the cooling efficiency of the coolant.

[0039] A smart stamping die failure monitoring system includes: a data acquisition layer, which uses a voltage sensor to detect voltage fluctuations in the stamping press, a temperature sensor to detect the coolant temperature of the stamping press, a vibration sensor to detect the vibration frequency of the stamping press, and a photoelectric encoder to collect stamping frequency data; for the steel strip to be processed, a magnetic flux detection device is set up to detect changes in the magnetic flux of the steel strip; the data collected by the data acquisition layer is timestamped and then transmitted to a data storage layer, which has a local SSD cache, before being transmitted to a cloud server database; a data processing layer extracts data from the data storage layer and performs data cleaning, feature extraction, and anomaly detection; then the data is transmitted to a model analysis layer; the model analysis layer performs failure assessment on the die, and after the assessment, the application control layer performs threshold alarm processing based on thresholds and optimizes the processing equipment parameters.

[0040] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0041] By introducing magnetic flux parameters to monitor mold failure, a failure model was constructed. Through setting a three-level early warning mechanism, the monitoring of mold failure was realized. Attached Figure Description Figure 1 This is a system workflow diagram for this system. Detailed Implementation

[0042] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0043] A method for monitoring failures in intelligent stamping dies:

[0044] Step 1, data acquisition: Collect key data by setting up sensors;

[0045] Individual data collection is performed on each stamping machine, including voltage, coolant temperature, machine vibration frequency, and stamping frequency. A magnetic flux detection device is installed for the steel strip to be processed. Data is timestamped synchronously to ensure accurate analysis.

[0046] Step 2: Data storage, storing the acquired data;

[0047] The data mentioned above is stored locally and then transmitted to the backend server.

[0048] Step 3: Data is transmitted to the backend server for preprocessing;

[0049] Remove outliers from the raw data, standardize and normalize the data, and extract feature values;

[0050] Step 4: Construct a parameter-based failure prediction model, input parameters into the model, and calculate the current wear state:

[0051] Starting from a multi-parameter linear fusion model, and combining failure weight allocation and normalization factors, a comprehensive failure scoring model is formed; each term represents the normalized influence degree of a failure influencing factor, constructing a comprehensive failure factor F. fail Its expression is:

[0052]

[0053] Among them: the weight coefficients w1 to w5 are optimized monthly (based on regression analysis of historical failure cases).

[0054] w1 to w5 are weighting coefficients, reflecting the degree of influence of each parameter on failure, with a value range of [0,1]. The optimal values ​​are: w1 = 0.15, w2 = 0.25, w3 = 0.2, w4 = 0.25, and w5 = 0.15.

[0055] V0 is the nominal voltage, also known as the rated voltage of the equipment;

[0056] ΔV is the absolute value of the voltage fluctuation;

[0057] T cool The temperature of the coolant generally fluctuates between 20 and 60 degrees Celsius, with ≤40 being the optimal state.

[0058] T ref This is the reference temperature, representing a normal temperature baseline value; the set value is 20 degrees Celsius.

[0059] T max The maximum permissible temperature is set to a safety limit of 60 degrees Celsius.

[0060] f vib The vibration frequency is typically 0–1000 Hz; values ​​below 300 Hz are considered ideal.

[0061] f0 is the normal vibration frequency, which is the frequency within the allowable voltage fluctuation range of the equipment. This frequency is the next normal operating frequency for the equipment; in this system, it is set to 200 Hz.

[0062] f max This is the upper limit of the vibration frequency, and the threshold for the system to mark an anomaly; the value is 1000 Hz.

[0063] N press This refers to the actual number of stamping operations, typically ranging from 0 to 120 times per minute. Ideally, it should be below 100 times per minute.

[0064] N0 is the base frequency, a steady-state value of the device, which is 90 times per minute.

[0065] N max The maximum allowed frequency is set at 200 times per minute;

[0066] Φ0 is the reference magnetic flux, and the normal magnetic flux value of the steel strip is 0.1~0.5wb. ΔΦ is the difference between the actual detected magnetic flux and the reference magnetic flux.

[0067] Step 5: Set alarm thresholds and perform real-time monitoring, automatically recording historical data for model optimization.

[0068] The system is configured with a tiered early warning mechanism:

[0069] Level 1 warning (F) fail ∈[0.5,0.7)): The system displays "Mold Status Offset," prompting the operator to pay attention;

[0070] Level II warning (F) fail ∈[0.7,0.9)): The system prompts "Mold status abnormal", and it is recommended to check the cooling system and voltage stability;

[0071] Level III Warning (F) fail ≥0.9): The system will automatically stop, prompting you to replace the mold or repair the equipment;

[0072] Step Six: Set up the feedback and optimization module, and adjust the process parameters according to the model output;

[0073] Adjust the parameters in step one to improve the mold's lifespan;

[0074] Including F fail When the value is greater than 0.6, reduce the stamping speed, thereby reducing the stamping frequency;

[0075] When T cool When the temperature is above 45 degrees Celsius, increase the cooling efficiency of the coolant.

[0076] A smart stamping die failure monitoring system includes: a data acquisition layer, which uses a voltage sensor to detect voltage fluctuations in the stamping press, a temperature sensor to detect the coolant temperature of the stamping press, a vibration sensor to detect the vibration frequency of the stamping press, and a photoelectric encoder to collect stamping frequency data; for the steel strip to be processed, a magnetic flux detection device is set up to detect changes in the magnetic flux of the steel strip; the data collected by the data acquisition layer is timestamped and then transmitted to a data storage layer, which has a local SSD cache, before being transmitted to a cloud server database; a data processing layer extracts data from the data storage layer and performs data cleaning, feature extraction, and anomaly detection; then the data is transmitted to a model analysis layer; the model analysis layer performs failure assessment on the die, and after the assessment, the application control layer performs threshold alarm processing based on thresholds and optimizes the processing equipment parameters.

Claims

1. A method for monitoring failure of intelligent stamping dies, characterized in that, The following steps are used: Step 1, data acquisition: Collect key data by setting up sensors; Step 2: Data storage, storing the acquired data; The data mentioned above is stored locally and then transmitted to the backend server. Step 3: Data is transmitted to the backend server for preprocessing; Remove outliers from the raw data, standardize and normalize the data, and extract feature values; Step 4: Construct a parameter-based failure prediction model, input parameters into the model, and calculate the current wear state: Starting from a multi-parameter linear fusion model, and combining failure weight allocation and normalization factors, a comprehensive failure scoring model is formed; each term represents the normalized influence degree of a failure influencing factor, constructing a comprehensive failure factor F. fail Its expression is: Where: w1~w5 are weighting coefficients, reflecting the degree of influence of each parameter on failure, and the value range is [0,1]; V0 is the nominal voltage, also known as the rated voltage of the equipment; ΔV is the absolute value of the voltage fluctuation; T cool The coolant temperature should fluctuate between 20 and 60 degrees Celsius, with ≤40 degrees Celsius being the optimal state. T ref This is the reference temperature, representing a normal temperature baseline value; the set value is 20 degrees Celsius. T max The maximum permissible temperature is set to a safety limit of 60 degrees Celsius. f vib The vibration frequency ranges from 0 to 1000 Hz; values ​​below 300 Hz represent an ideal state. f0 is the normal vibration frequency, which is the frequency at which the equipment operates within the allowable voltage fluctuation range; the frequency at which the equipment is in normal operation, and the value of this system is 200 Hz. f max This is the upper limit of the vibration frequency, and the threshold for the system to mark an anomaly; the value is 1000 Hz. N press The actual number of stamping cycles is 0 to 120 per minute; ideally, it is less than 100 per minute. N0 is the base frequency, a steady-state value of the device, which is 90 times per minute. N max The maximum allowed frequency is set at 200 times per minute; Φ0 is the reference magnetic flux; the normal magnetic flux value of the steel strip is 0.1~0.5wb. ΔΦ is the difference between the actual detected magnetic flux and the reference magnetic flux; Step 5: Set alarm thresholds and perform real-time monitoring; Step Six: Adjust process parameters according to the model; Adjust the parameters in step one to improve the mold's lifespan.

2. The intelligent stamping die failure monitoring method according to claim 1, characterized in that, In step one, individual data collection is performed on each stamping machine, including voltage, coolant temperature, machine vibration frequency, and stamping frequency. For the steel strip to be processed, a magnetic flux detection device is set up. Data is timestamped synchronously.

3. The intelligent stamping die failure monitoring method according to claim 1, characterized in that, The optimal values ​​for the weight coefficients w1 to w5 in step four are: w1 = 0.15, w2 = 0.25, w3 = 0.2, w4 = 0.25, and w5 = 0.

15.

4. The intelligent stamping die failure monitoring method according to claim 1, characterized in that, Step 5: Set up a tiered early warning mechanism: Level 1 warning (F) fail ∈[0.5,0.7)): The system displays "Mold status offset", prompting the operator to pay attention; Level II warning (F) fail ∈[0.7,0.9)): The system prompts "Mold status abnormal", it is recommended to check the cooling system and voltage stability; Level III Warning (F) fail ≥0.9): The system will automatically stop and prompt you to replace the mold or repair the equipment.

5. The intelligent stamping die failure monitoring method according to claim 1, characterized in that, Step Six: The system detected F fail When T > 0.6, reduce the stamping speed, thereby reducing the stamping frequency; when T cool When the temperature is above 45 degrees Celsius, increase the cooling efficiency of the coolant.

6. An intelligent stamping die failure monitoring system, characterized in that, Based on the steps of claims 1-5, the method includes: a data acquisition layer, which uses a voltage sensor to detect voltage fluctuations in the stamping machine, a temperature sensor to detect the coolant temperature of the stamping machine, a vibration sensor to detect the vibration frequency of the stamping machine, and a photoelectric encoder to acquire stamping frequency data; for the steel strip to be processed, a magnetic flux detection device is set to detect changes in the magnetic flux of the steel strip; the data collected by the data acquisition layer is timestamped and then transmitted to the data storage layer, which has a local SSD cache, and then transmitted to the cloud server database; the data processing layer extracts data from the data storage layer and performs data cleaning, feature extraction, and anomaly detection; then the data is transmitted to the model analysis layer; the model analysis layer performs failure assessment on the mold, and after the assessment, the application control layer performs threshold alarm processing based on thresholds and optimizes the processing equipment parameters.

Images