A system and method for early warning of flash fire failure in an aluminum foil electrochemical etching production line

Abstract

The application discloses a kind of aluminum foil electrochemical etching production line flash fire fault early warning system and method, comprising: synchronous acquisition electrode foil formation production line is the output voltage signal of two-way constant current source for foil positive and negative face power supply, corresponding obtain first voltage signal and second voltage signal;First voltage signal and second voltage signal are extracted to upper envelope line respectively, corresponding obtain first envelope signal and second envelope signal, and first envelope signal and second envelope signal are added in real time, obtain synthesis voltage signal;Characteristic extraction is carried out to synthesis voltage signal, obtains fluctuation characteristic parameter of voltage fluctuation;According to the relationship between fluctuation characteristic parameter and preset threshold, it is judged whether flash fire fault occurs and flash fire early warning is sent, and corresponding alarm information and early warning information are output.The application can effectively inhibit foil swing, tension fluctuation and other strong mechanical interference, significantly improve detection accuracy, avoid false alarm.

Description

Technical Field

[0001] This invention belongs to the field of electrode aluminum foil production and manufacturing technology, and particularly relates to a flash fire fault early warning system and method for aluminum foil electrochemical etching production line. Background Technology

[0002] Electrode aluminum foil is the core material of aluminum electrolytic capacitors. Its formation process involves applying constant current and high voltage to the foil in the electrolyte to form a dense oxide film on its surface, directly affecting the capacitor's key performance characteristics such as withstand voltage and lifespan. This process requires extremely high stability of parameters such as current, voltage, and temperature. However, in actual continuous production, due to interference from multiple factors such as uneven microstructure of the foil, local fluctuations in the electrolyte, changes in the contact resistance of the conductive roller, and high-speed oscillation of the foil, instantaneous breakdown discharge, or "flashover," can easily occur in local areas of the oxide film. Flashover causes irreversible damage to the oxide film, leading to a decrease in product withstand voltage and an increase in leakage current, which is a major process defect affecting the yield and performance level of electrode foil. Currently, electrode foil production lines generally use PLC-based macroscopic process parameter monitoring systems, which can only provide threshold alarms for steady-state parameters such as current, average voltage, and temperature, and are completely unable to capture millisecond-level transient flashover events. The industry still mainly relies on offline electrical performance testing for post-event evaluation, which is severely lagging and cannot achieve real-time intervention in the production process. Furthermore, technologies from other industrial fields, such as electrical fire monitoring, thermal imaging detection, and arc flash protection, cannot be directly applied to online detection of electrode foil flashover because their detection principles, application scenarios, and signal characteristics are completely different from those in this field.

[0003] In summary, existing technologies lack an effective solution for directly, online, in real-time, and with high accuracy detecting flashover faults during electrode foil formation. Current monitoring systems in this field can only handle macroscopic steady-state parameter drift, exhibiting blind spots for transient and random flashover events. Furthermore, direct application of arc, arc light, and temperature detection technologies from other fields, due to inherent differences in load characteristics, signal morphology, and interference environments, will result in severe false alarms or missed alarms, failing to meet the actual needs of production lines. Therefore, there is an urgent need for a non-intrusive solution specifically designed for the transient electrical characteristics of electrode foil flashover faults and the complex interference environment of production lines, capable of real-time online detection and early warning, to fill this technological gap and achieve proactive prevention and control of flashover faults. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a flashover fault early warning system and method for aluminum foil electrochemical etching production lines. It aims to achieve millisecond-level real-time, online direct detection of flashover faults during electrode foil production, filling a technological gap in this field. Through innovative signal processing methods, it effectively overcomes the influence of strong mechanical interference such as foil swaying and tension fluctuations on the detection signal, ensuring detection accuracy. It can not only determine flashover events that have already occurred, but also achieve probabilistic early warning of impending major flashovers by statistically analyzing small, predictive voltage jumps, transforming passive response into proactive prevention. It provides a complete, non-invasive solution that can be integrated into existing production lines, with low implementation cost and high reliability.

[0005] One of them, an early warning system for flashover faults in an aluminum foil electrochemical etching production line, includes:

[0006] The signal acquisition module is used to synchronously acquire the output voltage signals of two constant current sources that power the front and back sides of the foil in the electrode foil formation production line, and obtain the first voltage signal and the second voltage signal accordingly.

[0007] The signal preprocessing module is used to extract the upper envelope of the first voltage signal and the second voltage signal respectively, to obtain the first envelope signal and the second envelope signal respectively, and to add the first envelope signal and the second envelope signal in real time to obtain the synthesized voltage signal;

[0008] The data analysis module is used to extract features from the synthesized voltage signal, obtain fluctuation characteristic parameters that characterize voltage fluctuations, and determine whether a flashover fault and flashover warning have occurred based on the relationship between the fluctuation characteristic parameters and a preset threshold.

[0009] The output module is used to output alarm information and early warning information based on the judgment result of the data analysis module.

[0010] Preferably, the signal acquisition module includes:

[0011] A sampling line with a fuse is connected to the DC output ports of two constant current sources respectively, for directly extracting the first voltage signal and the second voltage signal;

[0012] An isolation amplifier, connected to the fuse-equipped sampling line, is used to electrically isolate and condition the first voltage signal and the second voltage signal, and outputs the isolated first voltage signal and the second voltage signal to the signal preprocessing module.

[0013] Preferably, the signal preprocessing module uses the moving window maximum value method to extract the upper envelope of the first voltage signal and the second voltage signal respectively;

[0014] The moving window maximum value method includes: sliding a window of a set length on the discrete voltage signal sequence, and taking the maximum value of all sampling points within the window as the envelope value of the corresponding point of the current window.

[0015] Preferably, the data analysis module includes:

[0016] The operating condition judgment unit is used to determine whether the production line is in a normalized operation state based on the average, peak and minimum values ​​of the synthesized voltage signal within a preset time window, combined with a preset operating condition threshold, and to start subsequent analysis when the state is determined to be normal.

[0017] Preferably, the data analysis module further includes:

[0018] The fault determination unit is used to calculate the fluctuation value and instantaneous rate of change of the synthesized voltage signal within a time window, and compare the fluctuation value with a first fluctuation threshold and the instantaneous rate of change with a first rate of change threshold. If the fluctuation value exceeds the first fluctuation threshold and the instantaneous rate of change exceeds the first rate of change threshold, a flashover fault is determined to have occurred.

[0019] The early warning counting unit is used to record voltage fluctuation events that meet the conditions of the fluctuation value exceeding the second fluctuation threshold and the instantaneous change rate exceeding the second change rate threshold but do not meet the judgment conditions of the fault judgment unit as early warning events, and count the number of early warning events within a sliding time window. When the number of early warning events exceeds the early warning counting threshold, a flash fire warning is triggered.

[0020] Wherein, the second fluctuation threshold is lower than the first fluctuation threshold, and the second rate of change threshold is lower than the first rate of change threshold.

[0021] This invention also provides a flashover fault early warning method for aluminum foil electrochemical etching production lines, comprising:

[0022] The output voltage signals of two constant current sources that supply power to the front and back of the foil in the electrode foil formation production line are simultaneously acquired to obtain the first voltage signal and the second voltage signal accordingly.

[0023] The upper envelope of the first voltage signal and the second voltage signal are extracted respectively to obtain the first envelope signal and the second envelope signal respectively. The first envelope signal and the second envelope signal are added in real time to obtain the synthesized voltage signal.

[0024] Feature extraction is performed on the synthesized voltage signal to obtain fluctuation characteristic parameters characterizing voltage fluctuations;

[0025] Based on the relationship between the fluctuation characteristic parameters and the preset threshold, it is determined whether a flashover fault has occurred and a flashover warning is issued, and corresponding alarm information and warning information are output.

[0026] Preferably, the process of obtaining the first envelope signal and the second envelope signal includes:

[0027] Using the moving window maximum value method, for the discrete sequences of the first voltage signal and the second voltage signal, the maximum value of all sampling points within the sliding time window is taken as the envelope value of the corresponding point in the current window, thereby obtaining the first envelope signal and the second envelope signal respectively.

[0028] Preferably, the process of obtaining the fluctuation characteristic parameters characterizing voltage fluctuations includes:

[0029] Within an analysis time window, the mean and peak values ​​of the synthesized voltage signal are calculated, and the fluctuation value is obtained based on the difference between the peak value and the mean value.

[0030] The instantaneous rate of change of voltage is calculated near the sampling point corresponding to the peak value.

[0031] Preferably, the process of determining that a flashover fault has occurred includes:

[0032] Set a first fluctuation threshold and a first rate of change threshold;

[0033] When the fluctuation value is greater than the first fluctuation threshold and the instantaneous rate of change is greater than the first rate of change threshold, a flashover fault is determined to have occurred.

[0034] Preferably, the process of determining the occurrence of a flash fire warning includes:

[0035] A second fluctuation threshold and a second rate of change threshold are set, wherein the second fluctuation threshold is less than a first fluctuation threshold and the second rate of change threshold is less than a first rate of change threshold;

[0036] A voltage fluctuation event that meets the conditions of fluctuation value greater than the second fluctuation threshold, instantaneous rate of change greater than the second rate of change threshold, and does not meet the criteria for flashover fault determination is recorded as an early warning event.

[0037] The number of warning events is counted within a sliding time window;

[0038] A flash fire warning is issued when the number of warning events exceeds the warning count threshold.

[0039] Compared with the prior art, the present invention has the following advantages and technical effects:

[0040] This invention achieves millisecond-level real-time online detection of flashover faults in electrode foil production lines for the first time by directly monitoring the output voltage waveform of the constant current source. This solves the detection blind spot problem of existing technologies that cannot capture instantaneous flashover events. Through a unique dual-channel signal envelope extraction and synthesis processing algorithm, it effectively suppresses strong mechanical interference such as foil swaying and tension fluctuations, significantly improving detection accuracy and avoiding false alarms. By statistically analyzing small-amplitude predictive voltage jumps, it can provide probabilistic early warnings before major flashovers occur, transforming traditional reactive remediation into proactive prevention, and significantly reducing the risk of batch quality accidents. At the same time, the entire solution does not require modification of core production line equipment, and has outstanding advantages such as low implementation cost and easy integration, providing reliable data support for electrode foil production process optimization and quality consistency control. Attached Figure Description

[0041] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0042] Figure 1 This is a schematic diagram of the system structure according to an embodiment of the present invention;

[0043] Figure 2 This is a schematic diagram of upper envelope signal extraction according to an embodiment of the present invention;

[0044] Figure 3 This is a schematic diagram of the flashover and the voltage fluctuation before the flashover in an embodiment of the present invention;

[0045] Figure 4 This is a graph showing the number of voltage jumps before a flashover in an embodiment of the present invention. Detailed Implementation

[0046] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0047] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0048] Example 1

[0049] The core concept of this embodiment lies in creatively using the fluctuation characteristics of the constant current source output voltage as a "fingerprint" signal for in-depth analysis of flashover faults. In the electrode foil formation circuit (constant current source → conductive ring → conductive roller → electrolyte → foil → electrolyte → conductive roller → conductive ring → constant current source), when the foil flashes, it is equivalent to instantaneously connecting a discharge channel with a high-resistance to low-resistance change in parallel in the circuit, which immediately causes transient fluctuations in the constant current source output voltage. This embodiment found that this fluctuation is not a simple step change, but a curve with a specific shape, and before a catastrophic flashover occurs, a series of small-amplitude voltage pre-flashover fluctuations with similar characteristics often appear.

[0050] Based on this discovery, the following technical solution is proposed in this embodiment:

[0051] Based on a non-invasive online monitoring system, its core function is to collect key electrical signals from the power supply circuit of the existing electrode foil forming production line and perform intelligent analysis. The system mainly consists of a signal acquisition module, a data processing module, and an output module. The data processing module includes a signal preprocessing module and a data analysis module.

[0052] This embodiment provides a flashover fault early warning system for an aluminum foil electrochemical etching production line, including:

[0053] The signal acquisition module is used to synchronously acquire the output voltage signals of two constant current sources that power the front and back sides of the foil in the electrode foil formation production line, and obtain the first voltage signal and the second voltage signal accordingly.

[0054] The signal preprocessing module is used to extract the upper envelope of the first voltage signal and the second voltage signal respectively, to obtain the first envelope signal and the second envelope signal respectively, and to add the first envelope signal and the second envelope signal in real time to obtain the synthesized voltage signal;

[0055] The data analysis module is used to extract features from the synthesized voltage signal, obtain fluctuation characteristic parameters that characterize voltage fluctuations, and determine whether a flashover fault has occurred and issue a flashover warning based on the relationship between the fluctuation characteristic parameters and a preset threshold.

[0056] The output module is used to output alarm and warning information based on the judgment results of the data analysis module.

[0057] This embodiment synchronously acquires the output voltage signals of two constant current sources supplying power to the front and back of the foil through a signal acquisition module, realizing millisecond-level real-time monitoring of the electrical signals of the formation circuit. This provides a direct and accurate data foundation for subsequent flashover feature extraction, thereby filling the technical gap in existing technologies that cannot capture transient flashover events online.

[0058] Furthermore, the signal acquisition module includes:

[0059] The sampling line with a fuse is connected to the DC output ports of the two constant current sources respectively, and is used to directly extract the first voltage signal and the second voltage signal;

[0060] An isolation amplifier, connected to a sample line with a fuse, is used to electrically isolate and condition the first voltage signal and the second voltage signal, and outputs the isolated first voltage signal and the second voltage signal to the signal preprocessing module.

[0061] Furthermore, the signal acquisition module in this embodiment includes a sampling line with a fuse and an isolation amplifier. The detection basis of this embodiment is monitoring the output voltage fluctuation of the constant current source. Therefore, it is necessary to acquire the DC output voltage of two independent constant current sources supplying power to the front and back sides of the foil with high precision and high real-time performance. Due to the formation process requirements, the two constant current sources cannot share a common ground; therefore, electrical isolation technology must be used to ensure that the signal is transmitted safely and without interference to the back-end processing unit.

[0062] Specifically, such as Figure 1 As shown, the implementation method and connection relationship of the signal acquisition module in this embodiment include:

[0063] Sampling lines A / B with fuses: Use dedicated cables with fuses to directly draw voltage signals V_A_raw and V_B_raw from the DC output ports (between positive and negative) of constant current source A and constant current source B. The fuses are used for overcurrent protection to ensure that a failure in the sampling circuit does not affect the safety of the main process circuit.

[0064] Isolation Amplifier: This is the key to the solution, used to solve the core problem of dual constant current sources not sharing a common ground. A mature industrial-grade multi-channel isolation amplifier is selected for voltage following, and internal magneto-electric or opto-coupling technology is used to achieve isolation between input, output, and power supply.

[0065] Connections: Sampling line A is connected to the input terminal of channel A of the isolation amplifier, and sampling line B is connected to the input terminal of channel B. The two output channels of the isolation amplifier are based on the same "clean" reference ground, outputting isolated signals V_A_iso and V_B_iso, which are connected to the data acquisition module of the data processing layer.

[0066] Signal conditioning: Filtering circuits (such as RC low-pass filters) can be integrated inside or outside the isolation amplifier to suppress high-frequency switching noise and power frequency interference, ensuring a "clean" voltage signal at the input end.

[0067] In this embodiment, the upper envelope of the two voltage signals is extracted and added in real time by the signal preprocessing module. The moving window maximum value method is used to effectively filter out the voltage common-mode noise caused by mechanical interference such as foil tension fluctuation and lateral swing. At the same time, the asymmetric voltage spike characteristics caused by the flashover event are enhanced, which significantly improves the detection signal-to-noise ratio and anti-interference ability, and avoids false alarms and missed alarms.

[0068] Furthermore, the signal preprocessing module uses the moving window maximum value method to extract the upper envelope of the first voltage signal and the second voltage signal respectively;

[0069] The moving window maximum value method involves sliding a window of a set length across a discrete voltage signal sequence and using the maximum value of all sampling points within the window as the envelope value of the corresponding point in the current window.

[0070] Furthermore, the calculation of the upper envelope involved in this embodiment is specifically performed using the moving window maximum value method.

[0071] In signal processing, the upper envelope is used to describe the local peak profile of signal oscillations. This embodiment employs the moving window maximum value method, a computationally efficient and real-time time-domain method suitable for processing quasi-DC voltage signals with mechanical interference, such as those from electrode foil production lines. The core idea is: for a discrete voltage signal sequence V[n] (where n is the sampling point index), a moving window of length L is defined. This window slides across the signal sequence, calculating the maximum value of all sampling points within the window each time, and using this maximum value as the envelope value corresponding to the center (or end) point of the window.

[0072] The formula is defined as follows:

[0073] Let the original discrete voltage signal be:

[0074] V_raw[n], n=1,2,3,...,N;

[0075] Where N is the total number of sampling points within an analysis time window.

[0076] Its envelope signal V_env[n] is calculated using the moving window maximum method:

[0077] V_env[n]=max(V_raw[nk],V_raw[n-k+1],...,V_raw[n],...,V_raw[n+m]);

[0078] For all n, where k and m define the front and back ranges of the window centered at the nth point, the total window length L = k + m + 1.

[0079] In practical programming or embedded implementations, a causal system (using only current and past data) is used to ensure real-time performance. In this case, the formula can be simplified to:

[0080] V_env[n]=max(V_raw[n-L+1],V_raw[n-L+2],...,V_raw[n]);

[0081] Here, L is the window length. This operation effectively filters out the gradual voltage drop caused by factors such as electrode foil tension relaxation, while preserving and highlighting rapid voltage rise events (flashover characteristics).

[0082] For the positive and negative voltage signals V_A_raw[n] and V_B_raw[n] in this embodiment, the upper envelope needs to be calculated separately to obtain:

[0083] V_A_env[n] = maximum value of the moving window(V_A_raw[n], L);

[0084] V_B_env[n] = maximum value of the moving window(V_B_raw[n], L);

[0085] Furthermore, the dual-channel signal synthesis in this embodiment involves real-time addition of the two signals after upper envelope processing. This is a key anti-interference step in this embodiment. Its purpose is to counteract the common-mode interference introduced by the reverse impedance change of the sensors on both sides caused by the lateral swing of the foil.

[0086] The synthesis formula is:

[0087] V_combined[n]=V_A_env[n]+V_B_env[n];

[0088] Wherein, V_combined[n] is the synthesized anti-interference voltage signal sequence.

[0089] In principle, foil oscillation causes an increase in impedance and voltage on one side of the circuit, while simultaneously decreasing impedance and voltage on the other side. After the two signals are processed by the upper envelope, the fluctuations caused by this oscillation are time-corresponding. By adding them together, this common-mode interference can theoretically be canceled out or significantly weakened. However, a real flashover event usually affects both circuits simultaneously, and its voltage fluctuation characteristics are preserved or even amplified in the synthesized signal.

[0090] This embodiment calculates the mean, peak, and minimum values ​​of the synthesized voltage signal using the operating condition judgment unit in the data analysis module and compares them with preset thresholds. This accurately identifies whether the production line is in a normalized operating state and initiates subsequent analysis only under normal operating conditions. This effectively eliminates invalid data from non-operating states such as shutdown, foil cutting, and abnormal shutdown, thereby improving the reliability of the system operation and the accuracy of the analysis results.

[0091] Furthermore, the data analysis module includes:

[0092] The operating condition judgment unit is used to determine whether the production line is in a normalized operation state based on the average, peak and minimum values ​​of the synthesized voltage signal within a preset time window, combined with a preset operating condition threshold, and to start subsequent analysis when the state is determined to be normal.

[0093] Furthermore, the data analysis module also includes:

[0094] The fault determination unit is used to calculate the fluctuation value and instantaneous rate of change of the synthesized voltage signal within the time window, and compare the fluctuation value with the first fluctuation threshold and the instantaneous rate of change with the first rate of change threshold. If the fluctuation value exceeds the first fluctuation threshold and the instantaneous rate of change exceeds the first rate of change threshold, a flashover fault is determined to have occurred.

[0095] The early warning counting unit is used to record voltage fluctuation events that meet the conditions of the fluctuation value exceeding the second fluctuation threshold and the instantaneous change rate exceeding the second change rate threshold but do not meet the judgment conditions of the fault judgment unit as early warning events, and count the number of early warning events within the sliding time window. When the number of early warning events exceeds the early warning counting threshold, a flash fire early warning is triggered.

[0096] Among them, the second fluctuation threshold is lower than the first fluctuation threshold, and the second rate of change threshold is lower than the first rate of change threshold.

[0097] This embodiment calculates the fluctuation value and instantaneous rate of change of the synthesized voltage signal through the fault determination unit in the data analysis module, and compares it with the first fluctuation threshold and the first rate of change threshold. When both exceed the threshold at the same time, a flashover fault is determined to have occurred. This realizes the accurate identification of flashover events by multiple parameters, overcomes the defect of single parameter determination being susceptible to process drift interference, and ensures high detection accuracy.

[0098] This embodiment sets a second fluctuation threshold and a second rate of change threshold below the fault threshold through the early warning counting unit in the data analysis module. Voltage fluctuation events that meet the conditions but do not reach the fault level are recorded as early warning events, and the number of early warning events is counted within a sliding time window. When the count exceeds the early warning threshold, a flashover early warning is triggered, realizing probabilistic prediction of major flashover faults. This allows operators to intervene and adjust process parameters in advance, transforming the traditional reactive handling into proactive prevention, and effectively reducing the risk of batch quality accidents.

[0099] This embodiment outputs alarm and / or warning information based on the judgment result through the output module, providing on-site operators with intuitive and timely fault prompts and risk warnings, facilitating rapid response and handling; at the same time, the flash fire event and warning event data recorded by the system can provide quantitative basis for subsequent process analysis and optimization, further improving the level of intelligent control of the production line.

[0100] Furthermore, this embodiment focuses on the calculation of statistical features within a time window. For the synthesized signal V_combined[n], its basic statistics need to be calculated within a fixed analysis time window (e.g., 6 seconds) for use in condition judgment and subsequent feature extraction.

[0101] Suppose that the sampling point index corresponding to this time window is from n_start to n_end, with a total of M = n_end - n_start + 1 points.

[0102] Mean (V_mean): Reflects the average voltage level within this window and serves as a benchmark for steady-state processes.

[0103] V_mean=(1 / M)*Σ_{n=n_start}^{n_end}V_combined[n];

[0104] This calculation formula is used to obtain the average level of the voltage signal within a time window.

[0105] Peak value (V_peak): This is the maximum voltage value within the window, used to capture the most significant fluctuations.

[0106] V_peak=max(V_combined[n_start],V_combined[n_start+1],...,V_combined[n_end])

[0107] Peak voltage is the maximum value of the voltage from the reference point to the highest point, and it is a key parameter describing the extreme values ​​of a signal.

[0108] Minimum value (V_min): The minimum voltage value within this window.

[0109] V_min=min(V_combined[n_start],V_combined[n_start+1],...,V_combined[n_end])

[0110] Fluctuation value (ΔV) calculation:

[0111] The fluctuation value is defined as the difference between the peak value and the mean value, which directly reflects the degree to which the voltage deviates from the steady state.

[0112] ΔV = V_peak - V_mean

[0113] This embodiment uses the difference between the peak value and the mean value to directly reflect the sudden deviation relative to the steady-state benchmark.

[0114] Furthermore, this embodiment also involves the calculation of the instantaneous rate of change (dV / dt). Specifically, the instantaneous rate of change is used to quantify the speed of voltage change and is a key parameter for distinguishing between rapid "flash" pulses and slow process drift. This rate of change needs to be calculated near the identified voltage peak point V_peak.

[0115] Let the index of the sampling point corresponding to V_peak be n_peak.

[0116] Difference approximation method: The derivative is approximated by calculating the difference between adjacent sampling points near the peak point;

[0117] dV / dt≈(V_combined[n_peak]-V_combined[n_peak-D]) / (D*Δt);

[0118] in:

[0119] D is the number of sampling points traced backward (for example, D=1 is used to calculate the rate of change between adjacent points, and a slightly larger D can smooth out noise).

[0120] Δt is the sampling interval time, Δt = 1 / sampling frequency.

[0121] To more robustly estimate the rate of change at the peak, we can also calculate the maximum value of the absolute values ​​of all differences within a small window (e.g., ±5 points) before and after the peak point:

[0122] dV / dt≈max(|V_combined[n_peak+j]-V_combined[n_peak+j-1]|) / Δt;

[0123] Where j takes values ​​in the range [-J, J].

[0124] The principle behind the formula is that the instantaneous rate of change is mathematically the derivative of voltage with respect to time, dV(t) / dt. In discrete systems, the differential is approximated using a difference. Its physical meaning is similar to the concept of rapid change embodied in the inductor voltage V=L*di / dt in circuit analysis, both emphasizing the importance of the rate of change. In fields such as synchronous control of power systems, the formula for calculating the rate of change of voltage often uses the difference between adjacent time points.

[0125] To further optimize the solution, the data processing module in this embodiment includes a signal preprocessing module and a data analysis module.

[0126] Specifically, the data processing module includes a data acquisition card and an industrial computer / embedded processor.

[0127] Function: Responsible for digitizing analog signals and running core algorithms to realize the extraction, judgment and early warning logic of flash fire characteristics.

[0128] Implementation methods and connection relationships:

[0129] Data acquisition card: A multi-channel synchronous sampling card or synchronous acquisition A / D module should be selected to ensure synchronous acquisition of two isolated voltage signals V_A_iso and V_B_iso. The sampling rate is set according to the frequency characteristics of the flash signal. In order to capture millisecond-level transients (50-250ms), the sampling rate should not be lower than 1kHz. In this embodiment, 3-10kHz is selected to ensure sufficient data points for accurate calculation of the rate of change (dV / dt).

[0130] Industrial computer / embedded processor: As the core computing unit, it communicates with the data acquisition card via a communication bus interface (SPI, PCIE / Ethernet, etc.) to acquire the digitized voltage sequences V_A[n] and V_B[n]. Simultaneously, the processor executes the fault warning method steps of this embodiment.

[0131] This embodiment achieves millisecond-level direct online detection of "flashover": for the first time, it directly correlates the most critical microscopic quality defect in electrode foil production—"flashover"—with an online, real-time measurable electrical signal (constant current source output voltage fluctuation), solving the long-standing problem of blind spots in online detection in this industry.

[0132] This embodiment possesses outstanding predictive and early warning capabilities: by statistically analyzing small, predictive fluctuations, it can issue early warnings before major flashovers occur, gaining valuable time for proactive process adjustments (such as speed reduction and contact point inspection). This is expected to transform post-event remediation into pre-event prevention, fundamentally reducing scrap rates. Actual measurements show that this early warning method achieves a flashover prediction accuracy of over 80%.

[0133] This embodiment boasts strong anti-interference capabilities and high detection accuracy: a specialized signal preprocessing algorithm was designed to address the unique mechanical interferences (tension fluctuations, foil oscillations) of the production line, effectively improving the signal-to-noise ratio. Combined with multi-parameter joint judgment logic, the false alarm rate can be significantly reduced. Actual measured flash fire detection accuracy reaches 100%.

[0134] This embodiment is low in implementation cost and easy to integrate: the solution only requires the addition of signal conditioning circuits and computing units, without the need to modify the core equipment of the production line, making it a cost-effective intelligent upgrade solution.

[0135] This embodiment provides data support for process optimization: the system records data such as the time, frequency, and amplitude of flashover and warning events, which provides a valuable data foundation for in-depth analysis of the relationship between process conditions (such as current density, electrolyte concentration, and speed) and flashover occurrence rate, and helps to continuously optimize the production process.

[0136] As a preferred implementation, this embodiment is applied to a high-voltage electrode foil formation production line with a foil width of 1.2 meters, a production speed of 20 meters / minute, and a constant current source output current of 3000A.

[0137] Hardware system composition:

[0138] Data acquisition unit: Simultaneously acquires two voltage signals with a sampling rate of 3125Hz to meet the requirement of capturing millisecond-level waveforms. It uses a Raspberry Pi as the processing core and runs the detection algorithm of this invention.

[0139] Algorithm implementation:

[0140] The data acquisition unit buffers data in 6-second blocks.

[0141] like Figure 2 As shown, the signal processing unit first performs 50Hz power frequency filtering on the two raw voltage data in each 6-second data block, then calculates the envelope of each data using the moving maximum method, and finally adds the envelope signals of the two data points together to obtain the synthesized signal V_combined.

[0142] like Figure 3 As shown, the operating condition judgment unit calculates the mean, peak, and minimum values ​​of V_combined over 6 seconds. A threshold is set: if the mean value > 16V or the minimum value < 8V, it is determined to be an abnormal operating condition, and the flashover analysis of this data block is skipped, waiting for the next data block.

[0143] Under normal operating conditions, the feature extraction unit calculates the mean V_mean of V_combined and finds its peak value V_peak. The fluctuation value ΔV = V_peak - V_mean is calculated. Near the peak point, the maximum absolute value of the voltage difference over 10 sampling points (corresponding to 2ms) is calculated as an approximation of the instantaneous rate of change dV / dt. Simultaneously, the time from when the voltage begins to rise above (V_mean + 0.1ΔV) to when it falls back below (V_mean + 0.1ΔV) is calculated as the duration T_duration.

[0144] Fault determination unit: Set fault thresholds ΔV_fault = 0.7V and dV / dt_fault = 1200V / s. If ΔV > ΔV_fault and dV / dt > dV / dt_fault, a flashover fault is determined to have occurred, the log is recorded, and an alarm is triggered. Simultaneously, T_duration is recorded.

[0145] Early warning counting unit: such as Figure 4 As shown, the warning threshold is set to ΔV_warn = 0.24V, and dV / dt_warn = 800V / s. If ΔV > ΔV_warn and dV / dt > dV / dt_warn, but the fault threshold is not reached, it is recorded as a warning event. The system maintains a warning event queue with a length of 12 hours (sliding window) and calculates the number of events in the queue, Count_warn, in real time.

[0146] Warning output unit: If Count_warn>10, a yellow warning signal is displayed on the human-machine interface (HMI), and the risk probability P=Count_warn / 10 (upper limit 1.0) is calculated. When P>0.8, a warning signal can be sent to the PLC to indicate a high risk.

[0147] Implementation results:

[0148] After a month of continuous testing on the production line, the system successfully detected all manually confirmed flashover events (39 in total), achieving a 100% detection accuracy. Of the 8 early warnings issued by the system, 7 occurred within 2 hours of the warning, resulting in a prediction accuracy of approximately 87.5%. This effectively helped operators intervene in advance, preventing several potential batch quality incidents.

[0149] Example 2

[0150] Based on the same inventive concept, this embodiment also provides a flashover fault early warning method for aluminum foil electrochemical etching production lines, including:

[0151] The output voltage signals of two constant current sources that supply power to the front and back of the foil in the electrode foil formation production line are simultaneously acquired to obtain the first voltage signal and the second voltage signal accordingly.

[0152] The upper envelope of the first voltage signal and the second voltage signal are extracted respectively to obtain the first envelope signal and the second envelope signal respectively. The first envelope signal and the second envelope signal are added in real time to obtain the composite voltage signal.

[0153] Feature extraction is performed on the synthesized voltage signal to obtain fluctuation characteristic parameters characterizing voltage fluctuations;

[0154] Based on the relationship between the fluctuation characteristic parameters and the preset threshold, it is determined whether a flashover fault has occurred and a flashover warning is issued, and corresponding alarm and warning information is output.

[0155] Furthermore, the process of obtaining the first envelope signal and the second envelope signal includes:

[0156] By employing the moving window maximum value method, for the discrete sequences of the first and second voltage signals, the maximum value of all sampling points within the sliding time window is taken as the envelope value of the corresponding point in the current window, thereby obtaining the first envelope signal and the second envelope signal respectively.

[0157] Furthermore, the process of obtaining the fluctuation characteristic parameters characterizing voltage fluctuations includes:

[0158] Within an analysis time window, the mean and peak values ​​of the synthesized voltage signal are calculated, and the fluctuation value is obtained based on the difference between the peak value and the mean value.

[0159] Calculate the instantaneous rate of change of voltage near the sampling point corresponding to the peak value.

[0160] Furthermore, the process of determining whether a flashover fault has occurred includes:

[0161] Set a first fluctuation threshold and a first rate of change threshold;

[0162] A flashover fault is determined to have occurred when the fluctuation value is greater than the first fluctuation threshold and the instantaneous rate of change is greater than the first rate of change threshold.

[0163] Furthermore, the process of determining the occurrence of a flash fire warning includes:

[0164] Set a second fluctuation threshold and a second rate of change threshold, wherein the second fluctuation threshold is less than the first fluctuation threshold and the second rate of change threshold is less than the first rate of change threshold;

[0165] A voltage fluctuation event that meets the conditions of fluctuation value greater than the second fluctuation threshold, instantaneous change rate greater than the second change rate threshold, and does not meet the criteria for flashover fault determination is recorded as a warning event.

[0166] Count the number of warning events within a sliding time window;

[0167] A flash fire warning is issued when the number of warning events exceeds the warning count threshold.

[0168] As a preferred implementation method, the solution of this embodiment specifically includes the following steps:

[0169] S1: Simultaneous sampling and preprocessing of dual-channel voltage signals;

[0170] On the electrode foil production line, two sets of constant current source circuits are connected to the front and back sides of the foil via conductive rollers. The system synchronously acquires the constant current source output voltage signals corresponding to the circuits on both sides, and the sampling frequency must be sufficient to capture millisecond-level transients (e.g., not less than 1kHz). Each analysis is performed in data blocks with a fixed time window (e.g., 6 seconds).

[0171] S2: Signal processing resistant to mechanical interference;

[0172] To eliminate measurement errors introduced by foil tension fluctuations (leading to a slow voltage drop) and lateral oscillations (leading to reverse impedance changes on both sides of the sensors) during the production process, this embodiment performs the following processing:

[0173] Upper envelope extraction: For each acquired raw voltage signal, its upper envelope is calculated (using the moving window maximum value method). This operation can effectively filter out interference caused by instantaneous voltage drops due to tension fluctuations, etc., and retain and highlight rapid voltage rise events (which is the main characteristic of flashover).

[0174] Dual-channel signal synthesis: The positive and negative voltage signals after upper envelope processing are added in real time. Since foil oscillation causes the impedance of one circuit to increase and the voltage to rise, while the impedance of the other circuit decreases and the voltage to fall, the common-mode oscillation interference can theoretically be canceled after the two signals are added, while the voltage fluctuation characteristics of a real flashover event (which usually affects both circuits simultaneously) will be enhanced.

[0175] S3: Production line condition assessment and data validity screening;

[0176] For the synthesized voltage signal, calculate its mean, peak, and minimum values ​​within a time window. Set thresholds based on process knowledge to determine whether the production line is in a normal formation operation state. For example:

[0177] If the average voltage is >16V or the minimum voltage is <8V, the production line is determined to be in an abnormal formation operation state, such as shutdown, startup, roller washing, foil cutting, or abnormal stoppage. In this state, subsequent flashover analysis and early warning will not be performed to avoid misjudgment. Otherwise, the production line is considered to be in a normal constant current formation state, and the core flashover characteristic analysis and judgment process will begin.

[0178] S4: Extraction and determination of core features of flash fire events;

[0179] For data determined to be in a normalized state, the following feature extraction and logical judgment are performed:

[0180] Calculate the baseline and fluctuation value: Calculate the average voltage (V_mean) within the time window as the baseline. Find the peak voltage (V_peak) within the window, and calculate the fluctuation value ΔV = V_peak - V_mean. This value reflects the degree to which the voltage deviates from the steady state.

[0181] Calculate the instantaneous rate of change: Calculate the instantaneous rate of change (dV / dt) near the voltage peak point, for example, by calculating the difference between several sampling points before and after the peak point. This parameter is used to distinguish between a fast "flashover" pulse and a slow process drift.

[0182] Determining if a flashover has occurred: Set a set of high threshold values, such as ΔV_threshold_fault > 0.7V and dV / dt_threshold_fault > 1200V / s. When both the extracted ΔV and dV / dt exceed these two thresholds, a flashover fault event is determined to have occurred. Simultaneously, the duration of this voltage fluctuation (from the rising edge to the time it falls back to near the reference level, typically ranging from 50ms to 250ms in actual measurements) can be recorded. This duration can be used to assist in assessing the severity of the flashover.

[0183] S5: Flash fire warning based on counting small voltage jumps;

[0184] A key innovation of this embodiment is that it can not only detect flashovers that have already occurred, but also predict flashovers that are about to occur. Long-term observation has revealed that in the period leading up to a significant flashover (as determined by S4), there are usually frequent small but rapid voltage jumps.

[0185] Warning event determination: Set a set of warning thresholds that are lower than the fault determination threshold, for example, ΔV_threshold_warning>0.24V and dV / dt_threshold_warning>800V / s. When the voltage fluctuation meets both of these conditions but does not reach the fault threshold, it is recorded as a warning event.

[0186] Sliding window counting and probability calculation: The system runs continuously and counts the warning events that occur within a relatively long sliding time window (e.g., 12 hours).

[0187] Warning Trigger: Based on actual production data, it was found that when the cumulative warning event count exceeds 10 within the window, the probability of a subsequent flashover fault as defined by S4 increases significantly. Therefore, Count_warning > 10 can be used as the condition for triggering a warning.

[0188] Risk Quantification: To further quantify the risk, this embodiment can calculate the probability of flashover P=min(Count_warning / 10,1.0). When the P value is close to or reaches 1, it indicates that flashover is very likely to occur in the near future, prompting operators or control systems to pay close attention or make process adjustments (such as checking the contact of the conductive rollers, adjusting the current or speed).

[0189] S6: System Output and Response;

[0190] The system outputs the following information in real time:

[0191] Real-time voltage waveform: Displays the processed composite voltage curve.

[0192] Fault alarm: When a flashover is detected, an audible and visual alarm is immediately triggered, and information such as the event time, fluctuation amplitude, and duration is recorded.

[0193] Warning prompt: When the number of warning events exceeds the threshold or the calculated probability P exceeds the set limit (e.g., 0.8), a warning prompt will be issued, indicating that there is an increased risk of flash fire on the production line.

[0194] Historical statistics: Provides historical statistical reports on flashover events and early warning events for process analysis and optimization.

[0195] The core of this embodiment lies in the real-time, online, and non-invasive detection of "flashover" faults during the electrode foil production process, and the provision of early warnings. Its key technical aspects lie in the precise capture of specific process signals, effective suppression of complex interference, and intelligent identification of fault characteristics. Specifically, this includes:

[0196] Direct monitoring of core process parameters: Unlike traditional offline, destructive testing methods (such as power frequency withstand voltage testing), the core of this embodiment is to directly monitor the output voltage waveform of the constant current source that provides energy to the electrode foil formation process. Because a "flashover" is essentially a momentary short circuit caused by localized breakdown of the electrode foil oxide film, this directly leads to a sudden change in load impedance, which is immediately reflected as a momentary fluctuation in the constant current source output voltage. This monitoring method has a direct signal source and a clear causal relationship, forming the basis for achieving millisecond-level real-time response.

[0197] A unique anti-interference signal processing mechanism: The electrode foil production line experiences strong mechanical vibrations and foil swaying interference, which can cause voltage signal fluctuations and are the main noise source for detecting "flashover" signals. This embodiment designs a combined anti-interference algorithm that uses dual-channel monitoring, upper envelope extraction, and signal summation. By monitoring two independent but related voltage signals and superimposing their upper envelopes, the algorithm effectively filters out low-frequency interference signals with periodicity and symmetry caused by mechanical tension fluctuations and foil swaying, thereby highlighting the asymmetric, sudden voltage spikes caused by "flashover".

[0198] The system employs a real-time analysis and early warning logic based on waveform characteristics. It not only detects the instantaneous absolute value of the voltage but, more importantly, analyzes its fluctuation value and rate of change (dV / dt). By setting reasonable thresholds, it can accurately capture the voltage drop characteristics that characterize "flashover." Furthermore, this embodiment achieves predictive early warning of impending severe flashover by statistically analyzing small, high-frequency predictive voltage fluctuations. This goes beyond simple fault alarms, enabling proactive perception and intervention of process conditions.

[0199] Non-invasive and system-integrated design: The entire solution requires no modification to existing core production equipment such as formation tanks and electrodes, nor does it require the introduction of complex external sensors such as optical and ion detectors. Functionality is achieved simply by acquiring voltage signals from the production line's electrical control cabinet and analyzing them through a deployed industrial computer or embedded system. This results in low implementation costs, high reliability, and easy integration into existing production control systems.

[0200] The flashover fault early warning method for aluminum foil electrochemical etching production line provided in this embodiment has all the advantages of the flashover fault early warning system for aluminum foil electrochemical etching production line provided in Embodiment 1.

[0201] Example 3

[0202] This embodiment also discloses a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described in Embodiment 1.

[0203] Example 4

[0204] This embodiment also discloses a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the method described in Embodiment 1.

[0205] Example 5

[0206] This embodiment also discloses a computer program product, including a computer program that, when executed by a processor, implements the steps of the method described in Embodiment 1.

[0207] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A flashover fault early warning system for an aluminum foil electrochemical etching production line, characterized in that, include: The signal acquisition module is used to synchronously acquire the output voltage signals of two constant current sources that power the front and back sides of the foil in the electrode foil formation production line, and obtain the first voltage signal and the second voltage signal accordingly. The signal preprocessing module is used to extract the upper envelope of the first voltage signal and the second voltage signal respectively, to obtain the first envelope signal and the second envelope signal respectively, and to add the first envelope signal and the second envelope signal in real time to obtain the synthesized voltage signal; The data analysis module is used to extract features from the synthesized voltage signal, obtain fluctuation characteristic parameters that characterize voltage fluctuations, and determine whether a flashover fault and flashover warning have occurred based on the relationship between the fluctuation characteristic parameters and a preset threshold. The output module is used to output alarm information and early warning information based on the judgment result of the data analysis module.

2. The system according to claim 1, characterized in that, The signal acquisition module includes: A sampling line with a fuse is connected to the DC output ports of two constant current sources respectively, for directly extracting the first voltage signal and the second voltage signal; An isolation amplifier, connected to the fuse-equipped sampling line, is used to electrically isolate and condition the first voltage signal and the second voltage signal, and outputs the isolated first voltage signal and the second voltage signal to the signal preprocessing module.

3. The system according to claim 1, characterized in that, The signal preprocessing module uses the moving window maximum value method to extract the upper envelope of the first voltage signal and the second voltage signal respectively. The moving window maximum value method includes: sliding a window of a set length on the discrete voltage signal sequence, and taking the maximum value of all sampling points within the window as the envelope value of the corresponding point of the current window.

4. The system according to claim 1, characterized in that, The data analysis module includes: The operating condition judgment unit is used to determine whether the production line is in a normalized operation state based on the average, peak and minimum values ​​of the synthesized voltage signal within a preset time window, combined with a preset operating condition threshold, and to start subsequent analysis when the state is determined to be normal.

5. The system according to claim 1, characterized in that, The data analysis module also includes: The fault determination unit is used to calculate the fluctuation value and instantaneous rate of change of the synthesized voltage signal within a time window, and compare the fluctuation value with a first fluctuation threshold and the instantaneous rate of change with a first rate of change threshold. If the fluctuation value exceeds the first fluctuation threshold and the instantaneous rate of change exceeds the first rate of change threshold, a flashover fault is determined to have occurred. The early warning counting unit is used to record voltage fluctuation events that meet the conditions of the fluctuation value exceeding the second fluctuation threshold and the instantaneous change rate exceeding the second change rate threshold but do not meet the judgment conditions of the fault judgment unit as early warning events, and count the number of early warning events within a sliding time window. When the number of early warning events exceeds the early warning counting threshold, a flash fire warning is triggered. Wherein, the second fluctuation threshold is lower than the first fluctuation threshold, and the second rate of change threshold is lower than the first rate of change threshold.

6. A method for early warning of flashover faults in an aluminum foil electrochemical etching production line, characterized in that, include: The output voltage signals of two constant current sources that supply power to the front and back of the foil in the electrode foil formation production line are simultaneously acquired to obtain the first voltage signal and the second voltage signal accordingly. The upper envelope of the first voltage signal and the second voltage signal are extracted respectively to obtain the first envelope signal and the second envelope signal respectively. The first envelope signal and the second envelope signal are added in real time to obtain the synthesized voltage signal. Feature extraction is performed on the synthesized voltage signal to obtain fluctuation characteristic parameters characterizing voltage fluctuations; Based on the relationship between the fluctuation characteristic parameters and the preset threshold, it is determined whether a flashover fault has occurred and a flashover warning is issued, and corresponding alarm information and warning information are output.

7. The method according to claim 6, characterized in that, The process of obtaining the first envelope signal and the second envelope signal includes: Using the moving window maximum value method, for the discrete sequences of the first voltage signal and the second voltage signal, the maximum value of all sampling points within the sliding time window is taken as the envelope value of the corresponding point in the current window, thereby obtaining the first envelope signal and the second envelope signal respectively.

8. The method according to claim 6, characterized in that, The process of obtaining the fluctuation characteristic parameters that characterize voltage fluctuations includes: Within an analysis time window, the mean and peak values ​​of the synthesized voltage signal are calculated, and the fluctuation value is obtained based on the difference between the peak value and the mean value. The instantaneous rate of change of voltage is calculated near the sampling point corresponding to the peak value.

9. The method according to claim 6, characterized in that, The process of determining if a flashover fault has occurred includes: Set a first fluctuation threshold and a first rate of change threshold; When the fluctuation value is greater than the first fluctuation threshold and the instantaneous rate of change is greater than the first rate of change threshold, a flashover fault is determined to have occurred.

10. The method according to claim 6, characterized in that, The process of determining when a flash fire warning has occurred includes: A second fluctuation threshold and a second rate of change threshold are set, wherein the second fluctuation threshold is less than a first fluctuation threshold and the second rate of change threshold is less than a first rate of change threshold; A voltage fluctuation event that meets the conditions of fluctuation value greater than the second fluctuation threshold, instantaneous rate of change greater than the second rate of change threshold, and does not meet the criteria for flashover fault determination is recorded as an early warning event. The number of warning events is counted within a sliding time window; A flash fire warning is issued when the number of warning events exceeds the warning count threshold.

Images