Water electrolysis hydrogen production fault diagnosis method based on empirical mode decomposition

By collecting the current signal of the electrolytic cell and analyzing it using wavelet denoising and empirical mode decomposition methods, the problem of troubleshooting caused by electromagnetic interference is solved, the cost is reduced and efficient health monitoring of the electrolytic water hydrogen production system is achieved.

CN118996533APending Publication Date: 2024-11-22SOUTHWEST PETROLEUM UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411152905.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-21
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

The existing fault diagnosis methods for electrolytic hydrogen production are susceptible to mutual interference of sensors in complex electromagnetic environments, resulting in unstable readings, and high-precision sensors and complex data processing systems increase equipment costs and maintenance costs.

Method used

By collecting the current signal of the electrolytic cell, the current signal is analyzed using wavelet denoising and empirical mode decomposition (EMD) methods, potential operational abnormalities are identified and fault warnings are issued, avoiding interference from different types of sensors and reducing the initial investment of the system and maintenance costs.

Benefits of technology

It realizes comprehensive status monitoring and management of electrolytic hydrogen production system, ensuring its safe, reliable and efficient operation, reducing costs, and being suitable for handling complex nonlinear and non-stationary signals.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118996533A_ABST
    Figure CN118996533A_ABST
Patent Text Reader

Abstract

The invention discloses a water electrolysis hydrogen production fault diagnosis method based on empirical mode decomposition, and the method comprises the following steps: analyzing current signals during faults of a water electrolysis cell, and finding out the incidence relation between state data; performing wavelet denoising on the current signal; processing the current signal by using an EMD method, and calculating an energy log value of a key IMF component; and the fault of the effluent electrolytic bath is judged. The analysis method (EMD) provided by the invention does not depend on any preset basis function, so that the analysis method is particularly suitable for processing non-linear and non-stationary signals and can be applied to analysis of current signals of the electrolytic cell.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of hydrogen production by electrolysis of water, and in particular to a fault diagnosis method for hydrogen production by electrolysis of water based on empirical mode decomposition. Background Art

[0002] Fault diagnosis technology for hydrogen production from water electrolysis is of great significance in ensuring system safety, optimizing operating efficiency and extending equipment life. First of all, health monitoring can detect the operating status of hydrogen production equipment from water electrolysis in real time and detect potential faults and abnormal conditions in a timely manner.

[0003] Existing fault diagnosis methods for hydrogen production by water electrolysis include monitoring the working voltage and current of the electrolyzer to ensure that the electrolysis process is carried out under optimal electrochemical conditions, or using various sensors to monitor the operating parameters of the electrolyzer. Temperature sensors monitor the temperature inside the electrolyzer and related components to prevent electrolyte evaporation and electrode material degradation caused by high temperature; pressure sensors monitor the pressure of hydrogen and oxygen to ensure that the pressure remains within a safe range during the generation and storage process; gas sensors are used to detect the purity of hydrogen and oxygen to ensure that the product gas meets the expected quality standards.

[0004] Chinese patent CN118064932A discloses an operation and maintenance method for water electrolysis hydrogen production equipment that is suitable for unattended operation. The method is based on the sensor architecture of the water electrolysis hydrogen production equipment. Sensors are set on the water electrolysis hydrogen production equipment to sense multi-parameter data, so as to obtain the sensed data of the water electrolysis hydrogen production equipment. A data analysis algorithm is set to perform fault warning analysis on the sensed data, and a health classification library is established. Based on the health classification library, the alarm information of the water electrolysis hydrogen production equipment is set to obtain operation and maintenance warning information.

[0005] However, different types of sensors installed in the same system may interfere with each other, especially in complex electromagnetic environments. Electromagnetic interference may cause unstable or erroneous sensor readings. The sensors themselves may also have varying degrees of impact on the operation of the electrolyzer. High-precision sensors and complex data processing systems may increase overall costs, especially in large-scale water electrolysis hydrogen production devices, where initial investment and maintenance costs are high.

[0006] Therefore, to solve these problems, the present invention only needs to collect the electrolytic cell current signal data, identify potential operating anomalies through advanced algorithm analysis, and provide real-time feedback; the fault warning and diagnosis system identifies abnormal patterns and issues warning signals by comparing historical data and real-time data, and the system automatically issues an alarm; the remote monitoring and management system allows operators to view the operating status of the water electrolysis hydrogen production system in real time on a remote platform via the Internet. Through this health monitoring method, the water electrolysis hydrogen production system can achieve comprehensive status monitoring and management to ensure its safe, reliable and efficient operation. Summary of the invention

[0007] The purpose of the present invention is to provide a method for analyzing the current signal of a water electrolyzer and ultimately realize fault diagnosis in different situations.

[0008] To achieve the above object, the present invention adopts the following technical solution:

[0009] Step 1: By analyzing the current signals of the water electrolyzer at various faults, the correlation between the state data is found, and the correlation analysis results are output to characterize the operating status of the system and provide a basis for the formulation of fault diagnosis strategies.

[0010] Step 2: Collect the current signal during the operation of the water electrolyzer through the current sensor.

[0011] Step 3: Perform wavelet denoising on the current signal.

[0012] Step 4: Use the EMD method to process the current signal, and the original signal is decomposed into several IMFs and a residual signal.

[0013] Step 5: Calculate the energy log value of the key IMF component and determine what kind of fault has occurred in the water electrolyzer.

[0014] In step 1, the water electrolyzer faults analyzed include reduced electrolyte flow rate, reduced electrolyte temperature, and reduced electrolyte concentration.

[0015] In step 3, the Symlet wavelet basis function is selected, the decomposition level is one, and the signal is subjected to discrete wavelet transform (DWT) to obtain detail coefficients and approximation coefficients. Approximation coefficient: low frequency part, representing the overall trend of the signal. Detail coefficient: high frequency part, representing the details of the signal.

[0016] Soft thresholding is used to remove the noise components in the detail coefficients, retain the main features of the signal, and use the coefficients after thresholding to perform inverse wavelet transform and reconstruct the denoised signal.

[0017] In step 4, the complex signal is decomposed into a series of intrinsic mode functions (IMFs) and a residual component using Empirical Mode Decomposition (EMD). The analysis process includes:

[0018] Find local maxima and local minima. Use spline interpolation to connect local maxima and local minima to construct upper and lower envelopes. Calculate the mean of the upper and lower envelopes. Extract the intrinsic mode function (IMF) and subtract the envelope mean from the signal to get a new signal.

[0019] If the new signal meets the definition conditions of IMF (i.e., the number of extreme points and the number of zero crossings of the signal are equal or differ by at most one, and the average value of the signal at any time is zero), it is considered to be an IMF; otherwise, continue to iterate the above steps until an IMF is obtained;

[0020] Subtract the obtained IMF from the original signal to obtain a residual signal. Repeat the above steps for the residual signal until the residual signal becomes monotonic or constant;

[0021] Finally, the original signal is decomposed into several IMFs and a residual signal.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] The present invention provides a water electrolyzer fault diagnosis and analysis method, which provides a basis for health monitoring of a complex operating system of electrolyzing water to produce hydrogen.

[0024] The present invention uses only current sensors to collect current signals, thus avoiding mutual interference caused by different types of sensors installed in the same system, especially in complex electromagnetic environments. Electromagnetic interference may cause unstable or erroneous sensor readings.

[0025] The initial investment and maintenance costs of the present invention are low.

[0026] The analysis method (EMD) proposed in the present invention does not rely on any preset basis function, and is therefore particularly suitable for processing nonlinear and non-stationary signals, and can be applied to the analysis of electrolytic cell current signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 A flow chart of a water electrolyzer fault diagnosis method based on empirical mode decomposition shown in an embodiment of the present invention;

[0028] Figure 2 The current signal when the electrolyte flow rate of the water electrolyzer is 40 mL / min shown in the embodiment of the present invention;

[0029] Figure 3 The current signal when the electrolyte flow rate of the water electrolyzer is 30 mL / min shown in the embodiment of the present invention;

[0030] Figure 4 The current signal when the electrolyte flow rate of the water electrolyzer is 20 mL / min shown in the embodiment of the present invention;

[0031] Figure 5 The empirical mode decomposition results of the electrical signals of various flow rates of the water electrolysis cell shown in the embodiment of the present invention. DETAILED DESCRIPTION

[0032] A method for diagnosing water electrolyzer faults based on empirical mode decomposition of the present invention is described in detail below in conjunction with the embodiments and drawings. The protection scope of the present invention includes but is not limited to the following description.

[0033] Figure 1 FIG. 1 is a flow chart of a method for diagnosing a water electrolyzer fault based on empirical mode decomposition according to the present invention. Figure 1 As shown, the water electrolyzer fault diagnosis method process includes the following steps:

[0034] Step 1: Conduct a simulation test on the abnormal temperature, flow rate and concentration of the electrolyte in the water electrolyzer. After the current signal is stable, use the current sensor to collect the electrical signal. The detailed experimental steps are as follows:

[0035] (1) The nickel plate was processed into a 30mm×10mm×2mm sample. The samples were polished step by step with water-abrasive paper. After polishing, the samples were washed with deionized water, washed with alcohol three times, dried with cold air, sealed in a storage container, and placed in a room temperature drying oven. The volume of the electrolytic cell is 150mL; the reference electrode is a saturated calomel electrode, and the electrolyte solution is a 30% NaOH solution; the auxiliary electrode is a platinum electrode, the platinum wire size is 10mm×5mm×0.1mm, and is encapsulated with epoxy resin. The diaphragm is selected as PPS diaphragm.

[0036] (2) The electrolyte flow rate of 40 mL / min, the electrolyte temperature of 65°C and the electrolyte concentration of 30% are regarded as the healthy working condition of the electrolytic cell. After the prepared NaOH solution is filled into the electrolytic cell, it is placed in a constant temperature water bath filled with an appropriate amount of clean water and heated to 65°C. The electrolyte flow rate is fixed at 40 mL / min using a water pump.

[0037] (3) To ensure the validity of the experimental data, the working electrode was first polarized at a voltage of 1.8 V for 60 min to remove the passivation film formed on the working surface in the air.

[0038] (4) Sampling frequency Fs = 10 Hz. After the electrolytic cell is operated under healthy working conditions for 24 hours, the temperature, concentration or flow rate is changed to obtain the current signals at each fault, such as Figure 2 , Figure 3 and Figure 4 shown.

[0039] Step 2: Perform wavelet denoising on the acquired current signal. By comparing the signal-to-noise ratio after denoising, select the Symlet wavelet basis function, decompose the level to one, and then perform discrete wavelet transform (DWT) on the signal to obtain detail coefficients and approximation coefficients. Use the coefficients after soft threshold processing to perform inverse wavelet transform and reconstruct the denoised signal.

[0040] Step 3: Perform EMD decomposition on the denoised signal, the first 8 IMF components of the healthy working condition signal and each fault signal, and calculate the log value of each IMF energy, such as Figure 5 shown.

[0041] Step 4: According to the results of EMD decomposition, it can be seen that the IMF6 characteristic energy can clearly distinguish the electrolytic cell flow rate fault. When the flow rate is 40mL / min (normal situation), the log value of its IMF energy is between -0.7 and 0; when the flow rate is 30mL / min (abnormal situation), the log value of its IMF energy is between -1.4 and -1; when the flow rate is 20mL / min (abnormal situation), the log value of its IMF energy is between -2.4 and -2. Therefore, the IMF6 characteristic energy can judge the electrolytic cell flow rate fault.

Claims

1. A fault diagnosis method for hydrogen production by water electrolysis based on empirical mode decomposition, characterized in that: The following steps are involved: Step 1: By analyzing the current signals of the water electrolyzer at various faults, the correlation between the state data is found, and the correlation analysis results are output to characterize the operating state of the system and provide a basis for the formulation of fault diagnosis strategies; Step 2: Collect the current signal during the operation of the water electrolyzer through the current sensor; Step 3: Perform wavelet denoising on the current signal; Step 4: Use the EMD method to process the current signal, and the original signal is decomposed into several IMFs and a residual signal; Step 5: Calculate the energy log value of the key IMF component and determine what kind of fault has occurred in the water electrolyzer.

2. The water electrolyzer fault diagnosis method based on empirical mode decomposition according to claim 1 is characterized in that: In step 1, the water electrolyzer faults analyzed include reduced electrolyte flow rate, reduced electrolyte temperature, and reduced electrolyte concentration.

3. The water electrolyzer fault diagnosis method based on empirical mode decomposition according to claim 1 is characterized in that: In step 3, the Symlet wavelet basis function is selected, the decomposition level is one, and the signal is subjected to discrete wavelet transform (DWT) to obtain detail coefficients and approximation coefficients.

4. The water electrolyzer fault diagnosis method based on empirical mode decomposition according to claim 1 is characterized in that: In step 3, soft thresholding is used to remove the noise components in the detail coefficients and retain the main features of the signal.

5. The water electrolyzer fault diagnosis method based on empirical mode decomposition according to claim 1 is characterized in that: In step 3, the coefficients after threshold processing are used to perform inverse wavelet transform and reconstruct the denoised signal.

6. The water electrolyzer fault diagnosis method based on empirical mode decomposition according to claim 1 is characterized in that: In step 4, the complex signal is decomposed into a series of intrinsic mode functions (IMFs) and a residual component using Empirical Mode Decomposition (EMD). The analysis process includes: S1: Find local maximum and local minimum, use spline interpolation to connect local maximum points and local minimum points, construct upper envelope and lower envelope, calculate the mean of upper envelope and lower envelope, extract intrinsic mode function (IMF), subtract the mean of envelope from the signal, and get a new signal; S2: If the new signal meets the definition conditions of IMF (i.e., the number of extreme points and the number of zero crossings of the signal are equal or differ by at most one, and the average value of the signal at any time is zero), it is considered to be an IMF; otherwise, continue to iterate the above steps until an IMF is obtained; S3: Subtract the obtained IMF from the original signal to obtain a residual signal, and repeat the above steps for the residual signal until the residual signal becomes monotonic or constant; S4: Finally, the original signal is decomposed into several IMFs and a residual signal.