Discharge test method and system for COB light bar

Abstract

This application relates to the field of discharge testing technology, specifically to a discharge testing method and system for COB lamp strips. The method includes: performing a discharge test on the COB lamp strip; collecting electrical variables of the COB lamp strip at various times during the test; performing wavelet decomposition on the electrical variables at all times within each time period to determine the noise interference level of the wavelet detail coefficients at each level, obtaining the first eigenvalue of each level of wavelet detail coefficients; using the first eigenvalue to correct the soft threshold in wavelet denoising at each level; determining the second eigenvalue for each time period; using the second eigenvalue to correct the soft threshold function in wavelet denoising; using the corrected wavelet denoising algorithm to denoise the electrical variables within each time period; performing anomaly detection on the denoised electrical variables; and obtaining the discharge test results of the COB lamp strip. This improves the accuracy of discharge testing of COB lamp strips.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of discharge test, in particular to a discharge test method and system for COB light bar. BACKGROUND

[0002] COB light bar is a kind of high-power integrated surface light source technology. By directly mounting LED chips on a mirror metal substrate to form a high-power density integrated light source, high light efficiency and uniform light distribution can be achieved. The electrical performance and stability of COB light bar directly determine its safety in long-term use. Discharge test is an important link to evaluate its electrical performance. By gradually increasing the working voltage beyond the rated working voltage of the light bar, the discharge voltage, current and other electrical variable parameters during the discharge process are monitored and analyzed after the internal insulating medium of the light bar is discharged, so as to evaluate the insulation performance and electrical performance of the light bar.

[0003] The existing discharge test method usually uses a discharge tester to test the light bar. During the working process of the light bar, the voltage, current and other parameters are monitored to identify whether there are abnormal peak fluctuations, partial discharge abnormalities and other fault phenomena caused by insulation defects in the COB light bar during the step-by-step voltage increase process, so as to test the electrical performance of the COB light bar. However, the instantaneous voltage fluctuation and partial discharge abnormality caused by the pulse noise generated by the discharge tester and the inevitable electrostatic interference in the environment will cause the electrical variable data to deviate and have false peak characteristics, resulting in a decrease in the accuracy of the obtained electrical variable data and errors in the discharge test results of the COB light bar. The traditional wavelet threshold denoising algorithm sets a fixed threshold to denoise the data. When facing the pulse noise and electrostatic interference with variable characteristics, it is difficult to effectively remove external interference, resulting in the inability to obtain accurate discharge test data to detect the stability of the electrical performance of the light bar. SUMMARY

[0004] In order to solve the above technical problems, the purpose of the present application is to provide a discharge test method and system for COB light bar, and the technical scheme adopted is as follows:

[0005] In the first aspect, the present application provides a discharge test method for COB light bar, which comprises the following steps:

[0006] Discharge test is performed on the COB light bar, and the electrical variables of the COB light bar at each time during the test process are collected, including voltage and current;

[0007] The discharge test duration of the COB light bar is evenly divided into time segments, and the electrical variables at all times in each time segment are wavelet decomposed to identify the maximum values in all wavelet detail coefficients in each layer and each local window; the mutation degree of each maximum value is analyzed to determine the noise impact weight of each maximum value, and the noise impact weight of each local window is further obtained;

[0008] By analyzing the distribution of noise impact weights in all local windows of each layer, the noise interference level of wavelet detail coefficients in each layer is determined. Combined with the disorder level of wavelet detail coefficients in each layer, the first eigenvalue of wavelet detail coefficients in each layer is obtained. The soft threshold in wavelet denoising of each layer is then corrected using the first eigenvalue.

[0009] The instantaneous phase changes of voltage and current at adjacent moments within each time period are determined, the differences in the voltage and current changes within each time period are analyzed, and the second characteristic value of each time period is determined; the second characteristic value is used to correct the soft threshold function in wavelet denoising.

[0010] A modified wavelet denoising algorithm was used to denoise the electrical variables in each time period, and anomaly detection was performed on the denoised electrical variables to obtain the discharge test results of the COB light strip.

[0011] In one embodiment, determining the noise impulse weights for each maxima includes:

[0012] Calculate the mean of the absolute values ​​of all wavelet detail coefficients remaining after removing the maxima within each local window of each layer. Calculate the difference between each maxima and the mean, and denote it as the first difference. The noise impact weight of each maxima is the proportion of the first difference of each maxima in the first difference of all maxima within its local window.

[0013] In one embodiment, obtaining the noise impact weights of each local window includes:

[0014] The noise impact weights of all maximum values ​​within each local window of each layer are sorted in descending order, and the fusion value of the first preset number of noise impact weights is used as the noise impact weight of each local window.

[0015] In one embodiment, determining the noise interference level of each wavelet detail coefficient includes:

[0016] By integrating the mean and dispersion of the noise impact weights of all local windows in each layer, the noise interference degree of the wavelet detail coefficients in each layer is obtained.

[0017] In one embodiment, the first feature value is the sum of the normalized value of the energy entropy of all wavelet detail coefficients in each layer and the normalized value of the noise interference degree of the wavelet detail coefficients in each layer.

[0018] In one embodiment, the soft threshold in wavelet denoising of each layer is corrected using the first feature value, and the expression is:

[0019] In the formula, The corrected soft threshold is used in the j-th layer wavelet denoising. Let be the first eigenvalue of the wavelet detail coefficients at the j-th level. This is a fixed threshold in wavelet denoising.

[0020] In one embodiment, the second characteristic value is the difference between the sum of the instantaneous phase changes of the voltage at all adjacent moments within each time period and the sum of the instantaneous phase changes of the current at all adjacent moments.

[0021] In one embodiment, the soft threshold function in wavelet denoising, which utilizes the second eigenvalue to correct the soft threshold function, is expressed as:

[0022] In the formula, Let be the second characteristic value of the nth time interval, where Sig() is the Sigmoid function and sgn() is the sign function. These are the wavelet detail coefficients of the j-th level and k-th wavelet after wavelet decomposition. represents the k-th new wavelet detail coefficient of the j-th layer after soft thresholding.

[0023] In one embodiment, obtaining the discharge test results of the COB light strip includes:

[0024] If any outlier is found among all the electrical variables after noise reduction, the COB light strip is determined to have an electrical safety risk; otherwise, the COB light strip is determined not to have an electrical safety risk.

[0025] Secondly, embodiments of this application also provide a discharge testing system for COB light strips, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of any of the methods described above.

[0026] This application has at least the following beneficial effects:

[0027] This application, through wavelet decomposition and analysis of the detail coefficients of electrical variables, can accurately identify abnormal fluctuations in electrical variables within each time period. In particular, the analysis of maxima in the detail coefficients can capture instantaneous noise impacts. By determining the noise impact weight of each local window, the noise interference degree of each layer of wavelet detail coefficients is obtained, accurately judging the degree of noise influence within each time period. This enhances the ability to identify noise impacts during testing and helps to more accurately reflect the actual performance of COB lamps during the discharge process. Secondly, the first eigenvalue of each layer of wavelet detail coefficients is obtained. Using the first eigenvalue to correct the soft threshold in wavelet denoising effectively solves the problem of inappropriate threshold selection in traditional methods, avoiding over-denoising or under-denoising, and ensuring the true reflection of COB lamp test data. Furthermore, by combining the instantaneous phase changes of voltage and current, the soft threshold function is corrected using the second eigenvalue, making the wavelet denoising algorithm more adaptable to processing electrical variables in different time periods. This approach addresses fluctuations in different electrical environments, enhancing the reliability and stability of the entire testing process and ensuring the stability and effectiveness of COB LED strip discharge test results in real-world applications. Finally, a modified wavelet denoising algorithm is used to denoise the electrical variables across different time periods. Anomaly detection is then performed on the denoised electrical variables to obtain the COB LED strip discharge test results. This avoids the loss of detailed information during denoising, ensuring the accuracy and completeness of the test results after denoising. This guarantees a comprehensive analysis of the COB LED strip performance, thereby improving the sensitivity of anomaly detection, reducing the impact of pulse noise and unavoidable electrostatic interference during testing, and enhancing the accuracy of the discharge test data to obtain more precise COB LED strip discharge test results. Attached Figure Description

[0028] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1 A flowchart illustrating the steps of a discharge testing method for a COB lamp strip, provided in one embodiment of this application;

[0030] Figure 2 The flowchart shows the soft threshold function correction process for the wavelet denoising algorithm. Detailed Implementation

[0031] To further illustrate the technical means and effects adopted by this application to achieve the intended inventive purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a discharge testing method and system for COB lamp strips proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0032] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0033] The following description, in conjunction with the accompanying drawings, details the specific scheme of the discharge testing method and system for COB lamp strips provided in this application.

[0034] Please see Figure 1 The diagram illustrates a flowchart of a discharge testing method for a COB lamp strip according to an embodiment of this application. The method includes the following steps:

[0035] S1, perform a discharge test on the COB lamp strip and collect electrical variables of the COB lamp strip at various times during the test, including voltage and current.

[0036] The COB LED strip is connected to a discharge testing instrument for discharge testing. The range of the test input voltage parameter is set; its value can be determined by the implementer based on the COB LED strip in the implementation scenario, without special restrictions. In this embodiment, the input voltage range is 50% to 200% of the COB LED strip's rated operating voltage. This embodiment sets the discharge test duration for the COB LED strip to 30 seconds; the implementer can set this duration according to actual conditions.

[0037] In the discharge test of the COB light strip, the electrical variable data of the COB light strip at various times are monitored. In this embodiment, the electrical variable data includes the operating voltage and operating current of the COB light strip. The data acquisition frequency is 2kHz, which can be set by the implementer according to the actual situation. It should be noted that the operating voltage and operating current of the COB light strip are acquired synchronously.

[0038] S2, the discharge test duration of the COB light strip is divided into various time periods. Wavelet decomposition is performed on the electrical variables at all times within each time period to identify the maxima in all wavelet detail coefficients within each local window of each layer. The degree of abrupt change of each maxima is analyzed to determine the noise impact weight of each maxima, and then the noise impact weight of each local window is obtained.

[0039] Given that the pulse noise generated by the discharge tester during operation and the electrostatic interference in the environment can cause instantaneous voltage fluctuations and partial discharge anomalies, resulting in deviations and false peaks in electrical variable data, and that noise interference has a high degree of randomness, traditional wavelet threshold denoising algorithms denoise the data based on a set fixed threshold. However, the input voltage continuously and dynamically changes during the discharge test, and the discharge test data also changes under different input voltage ranges. As a result, in this dynamic scenario, traditional wavelet threshold denoising algorithms are difficult to effectively remove external interference when faced with pulse noise and electrostatic interference with varying characteristics, ultimately affecting the accuracy of COB light bar discharge testing.

[0040] Based on the above analysis, this embodiment analyzes the electrical variable data collected during the discharge test of COB LED strips, taking the operating voltage as an example. During the discharge test of the COB LED strips, the electrical performance of the strips is measured by applying input voltages of different intensities. Considering that the input voltage gradually increases over time during the entire discharge test, the discharge test data will also change accordingly. To avoid the influence of the trend change in the discharge test data on noise reduction, this embodiment divides the entire discharge test duration into N time periods. In this embodiment, N=15. Implementers can set this value according to their actual situation; this embodiment does not impose any restrictions on this.

[0041] For the operating voltage of the COB light strip at all times within each time period, the db wavelet basis function is used to perform multi-level wavelet decomposition. The number of wavelet decomposition levels is J, and the size of J can be set by the implementer according to the implementation scenario without any special restrictions. The larger the value, the more refined the decomposition of the original signal. For signals with higher acquisition frequencies and more sampling points, the amount of information contained is often greater. Therefore, a larger decomposition level can be set to obtain more detailed changes in the signal. In this embodiment, J=3, and the wavelet decomposition yields the [number of]th [decomposition levels]. Layer Each wavelet detail coefficient is represented as follows: , representing the high-frequency components of different frequency bands.

[0042] In discharge testing, normal voltage and current data should exhibit stable and regularly changing power frequency voltage and current waveforms. However, due to their inherent uncertainty, impulse noise and electrostatic interference typically cause transient and severe impacts on electrical variable signal data, affecting the originally normal electrical variable data signal waveform. In terms of amplitude, this interference causes sudden spikes in the electrical variable signal waveform, resulting in a significant deviation of its amplitude from the baseline waveform. This phenomenon is similar to the partial discharge caused by insulation abnormalities in LED strips during discharge testing, making it easily misdiagnosed as a fault.

[0043] After wavelet decomposition, the aforementioned impulse noise and electrostatic interference exhibit a more prominent large-amplitude coefficient feature in the wavelet detail coefficients than normal electrical variable data. Unlike the anomalies caused by electrical performance faults, sudden impulse noise and electrostatic interference cause the wavelet detail coefficients of electrical variable data to exhibit a sparse feature of random distribution, while electrical performance faults will cause the signal characteristics to be abnormal throughout the continuous time period. There is a certain difference between the two.

[0044] Based on the above analysis, for the wavelet detail coefficients of each layer after wavelet decomposition, this embodiment divides all wavelet detail coefficients of each layer into S local windows of equal length, that is, the number of wavelet detail coefficients contained in each local window is equal. In this embodiment, S=20. Implementers can set it according to the actual situation. This embodiment does not limit it.

[0045] Considering that the abnormal amplitude of wavelet detail coefficients caused by impulse noise and electrostatic interference is greater than that of wavelet detail coefficients of stable signals, this embodiment searches for local maxima points in each local window as suspected noise impact points.

[0046] Specifically, take the j-th layer and the... The k-th wavelet detail coefficient in a local window If satisfied ,and Then the wavelet detail coefficients Recorded as the number A local maximum point within a local window, where... For the j-th layer The (k+1)th wavelet detail coefficient in a local window For the j-th layer The wavelet detail coefficients of the (k-1)th wavelet within a local window. Let the wavelet detail coefficient of the j-th layer be... The total number of local windows obtained We identify several maxima and determine the noise impact weight for each maxima. Then, we obtain the noise impact weight for each local window. Specifically:

[0047] Calculate the mean of the absolute values ​​of all wavelet detail coefficients remaining after removing the maxima within each local window of each layer, calculate the difference between each maxima and the mean, and denote it as the first difference. Obtain the noise impact weight of each maxima based on the first difference.

[0048] It should be noted that the difference represents the degree of difference between two variables, and can be calculated using methods such as difference, absolute value of difference, square of difference, ratio, etc. This embodiment does not limit this.

[0049] In this embodiment, the expression for the noise impulse weight of each maxima is:

[0050] In the formula, Indicates the first Layer The first in the local window The noise impact weight of the maximum value, Indicates the first Layer The first in the local window A maximum value, Indicates the first Layer The mean of the absolute values ​​of all wavelet detail coefficients remaining in a local window after removing the maxima. For the first Layer The first in the local window The nth local maximum, M is the nth local maximum. Layer The number of maxima within each local window. This is denoted as the first difference.

[0051] Noise impact weight Indicates the first The difference between the magnitude of the first maximum and the magnitudes of other normal wavelet detail coefficients can characterize the magnitude of the first maximum. The probability that each maximum point is a noise impact point is calculated, and the maximum value is more clearly seen after normalization. The proportion of all suspected noise impact points within this local window makes data differences more apparent. Noise impact weight. The larger the value, the better. The greater the difference between it and other wavelet detail coefficients, and the more prominent it is among all suspected noise impact points, the more likely it is due to abnormal intensity energy characteristics caused by interference.

[0052] The noise impact weights of all maxima within each local window of each layer are sorted in descending order, and the fused value of the first preset number of noise impact weights is used as the noise impact weight of each local window. The preset number can be set by the implementer according to actual conditions; this embodiment does not impose any restrictions on it.

[0053] It should be noted that fusion means combining multiple variables, which can be done by addition, multiplication, a combination of addition and multiplication, or by taking the average.

[0054] In this embodiment, the average of the first three noise impact weights arranged in descending order within each local window is taken as the noise impact weight of each local window.

[0055] S3. By analyzing the distribution of noise impact weights in all local windows of each layer, the noise interference degree of wavelet detail coefficients in each layer is determined. Combined with the disorder degree of wavelet detail coefficients in each layer, the first eigenvalue of wavelet detail coefficients in each layer is obtained. The soft threshold in wavelet denoising of each layer is corrected using the first eigenvalue.

[0056] Furthermore, the noise impact weights of all local windows in each layer are calculated to determine the noise interference level of the wavelet detail coefficients in each layer. Specifically, the fusion result of the mean and dispersion of the noise impact weights of all local windows in each layer is used as the noise interference level of the wavelet detail coefficients in each layer. In this embodiment, the sum of the mean and dispersion of the noise impact weights of all local windows in each layer is used as the noise interference level of the wavelet detail coefficients in each layer.

[0057] It should be noted that the degree of dispersion can be calculated using methods such as variance, standard deviation, and coefficient of variation. In this embodiment, variance is used as the method for calculating the degree of dispersion.

[0058] The greater the noise interference, the more suspected noise impact interference there is in the wavelet detail coefficients of the corresponding layer. However, it is not possible to accurately determine whether it is noise impact interference based solely on the amplitude of the wavelet detail coefficients, as it may also be a fault characteristic caused by abnormal electrical performance of the light strip.

[0059] Furthermore, consider the first The complexity of the energy distribution is analyzed by examining the entropy characteristics of the wavelet detail coefficients at each layer. The calculation of the first... Energy entropy of all wavelet detail coefficients in the layer The method for calculating the energy entropy of wavelet detail coefficients is existing technology, and its specific process will not be elaborated here. Energy Entropy The larger the value, the higher the value. The more discrete the energy distribution of the wavelet detail coefficients, the more complex the characteristic patterns within those coefficients, and the more likely they are to contain significant noise interference. Energy entropy The smaller the value, the higher the value. The more stable the energy distribution of the layer wavelet detail coefficients, the more characteristic it is of normal power frequency power or fault characteristics.

[0060] In summary, this embodiment calculates the first... First eigenvalue of layer wavelet detail coefficients The specific expression is:

[0061] In the formula, Let be the noise interference level of the wavelet detail coefficients at the j-th level. Let the energy entropy be the energy of all wavelet detail coefficients at the j-th level. This is the Sigmoid function.

[0062] Among them, the first eigenvalue The larger the value, the stronger the noise impact interference in the wavelet detail coefficients of that layer. Therefore, the wavelet detail coefficients of that layer should be subjected to higher denoising intensity to suppress the influence of impulse noise and electrostatic interference on the signal. The soft threshold in wavelet denoising of each layer is corrected using the first eigenvalue, specifically expressed as:

[0063] In the formula, The corrected soft threshold is used in the j-th layer wavelet denoising. Let be the first eigenvalue of the wavelet detail coefficients at the j-th level. The fixed threshold is used in wavelet denoising, and the determination of the fixed threshold is a well-known existing technique.

[0064] By determining the modified soft threshold in wavelet denoising as described above, the threshold determination in wavelet denoising can be adaptively adjusted in combination with the noise interference level of electrical variables in COB lamp strip discharge test, thereby improving the accuracy of threshold determination.

[0065] S4, determine the instantaneous phase changes of voltage and current at adjacent moments within each time period, analyze the differences in the changes of voltage and current within each time period, and determine the second characteristic value of each time period; use the second characteristic value to correct the soft threshold function in wavelet denoising.

[0066] During the discharge test, the input voltage of the COB LED strip is a regulated AC voltage during each time period. Therefore, the phase of the discharge voltage during the discharge test should also be stable within a certain range during this period, maintaining an ideal sine waveform. The discharge current signal is affected by the electrical characteristics of the LED strip itself. The current phase and voltage phase will show a certain lag or advance. That is, the instantaneous phase at each moment may be asynchronous, but the instantaneous phase difference should remain consistent over a period of time.

[0067] In the discharge test of COB LED strips, the instantaneous phase of the operating voltage and current data of the COB LED strips in each time period is obtained by performing a Hilbert transform. The Hilbert transform is a well-known technique, and its specific process will not be elaborated further. Based on the above analysis, the second characteristic value for each time period is determined, and its specific expression is as follows:

[0068] ;in, This is the second characteristic value in the nth time period during the COB lamp strip discharge test. , They represent the nth time interval and the nth time interval respectively. The, the The instantaneous phase of the operating voltage. , They represent the nth time interval and the nth time interval respectively. The, the The instantaneous phase of the operating current. Indicates the first Number of sampling points in each time period. It represents the instantaneous phase change of the voltage between adjacent moments. It represents the change in the instantaneous phase of the current between adjacent moments.

[0069] Second eigenvalue Indicates the first The phase difference between the operating voltage and operating current over a given time period can measure the phase consistency of the voltage and current data within that time period. If A value close to 0 indicates a smaller phase difference, meaning the discharge voltage and discharge current data maintain phase consistency. If... The greater the deviation of the value from 0, the more pulse noise and electrostatic interference there is, and the cross-channel phase consistency of the data is disrupted.

[0070] Furthermore, the soft threshold function in wavelet denoising of the COB light strip's operating voltage in the nth time period is corrected using the second eigenvalue of the nth time period. The expression is as follows:

[0071] In the formula, Let be the second characteristic value of the nth time interval, where Sig() is the Sigmoid function and sgn() is the sign function. These are the wavelet detail coefficients of the j-th level and k-th wavelet after wavelet decomposition. These are the k-th new wavelet detail coefficients at the j-th layer after soft thresholding. The flowchart for the soft thresholding function correction in the wavelet denoising algorithm is shown below. Figure 2 As shown.

[0072] It should be understood that when The closer to 0 The closer the value is to 0.5, the higher the phase consistency between the discharge voltage and discharge current, allowing for noise reduction using a standard soft threshold function; when... The further away from 0 The closer the value is to 1, the more the phase consistency between the discharge voltage and the discharge current is disrupted, thus requiring an improvement in the denoising effect of the soft threshold function.

[0073] A modified wavelet denoising algorithm was used to denoise the operating voltage of the COB lamp strip during each time period of the discharge test. Similarly, the operating current of the COB lamp strip during each time period of the discharge test was denoised using the same method as for the operating voltage, thus obtaining the denoised operating voltage and operating current for each time period of the COB lamp strip discharge test. The wavelet denoising algorithm is a well-known existing technology, and its specific process is not described in detail.

[0074] It should be noted that when performing wavelet denoising on the operating voltage and current of the COB LED strips in each time period, the second eigenvalue remains the same for each time period, as described above. .

[0075] S5 uses a modified wavelet denoising algorithm to denoise the electrical variables in each time period, performs anomaly detection on the denoised electrical variables, and obtains the discharge test results of the COB light strip.

[0076] The discharge test results of COB lamp strips were analyzed using the denoised discharge test electrical variable data. Specifically, taking the denoised operating voltage of the COB lamp strip in the discharge test as an example, a 3... The standard performs anomaly detection on the COB LED strip's denoised operating voltage at all times during the discharge test, specifically when the denoised operating voltage is within a certain range. If the data falls within a certain range, the denoised operating voltage is considered normal; otherwise, it is considered abnormal. 3 The guidelines are existing and well-known technologies. Implementers may choose other existing feasible anomaly detection algorithms at their own discretion. This embodiment does not impose any restrictions on this.

[0077] The same anomaly detection algorithm as the denoised operating voltage is used to detect anomalies in the denoised operating current. For the electrical variable data collected during the COB light strip discharge test, if any anomaly is found in any of the denoised electrical variables, the COB light strip is determined to have an electrical safety risk; otherwise, it is determined that the COB light strip does not have an electrical safety risk.

[0078] Based on the same inventive concept as the above method, this application embodiment also provides a discharge testing system for COB light strips, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described discharge testing methods for COB light strips.

[0079] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, specific embodiments of this specification have been described above. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0080] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0081] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the principles of this application should be included within the protection scope of this application.

Claims

1. A discharge test method for COB lamp strips, characterized in that, The method includes the following steps: A discharge test was performed on the COB LED strip, and the electrical variables of the COB LED strip at various moments during the test were collected, including voltage and current. The discharge test duration of the COB light strip is divided into time periods. Wavelet decomposition is performed on the electrical variables at all times within each time period to identify the maxima in all wavelet detail coefficients in each local window of each layer. The degree of change of each maxima is analyzed to determine the noise impact weight of each maxima, and then the noise impact weight of each local window is obtained. By analyzing the distribution of noise impact weights in all local windows of each layer, the noise interference level of wavelet detail coefficients in each layer is determined. Combined with the disorder level of wavelet detail coefficients in each layer, the first eigenvalue of wavelet detail coefficients in each layer is obtained. The soft threshold in wavelet denoising of each layer is then corrected using the first eigenvalue. The instantaneous phase changes of voltage and current at adjacent moments within each time period are determined, the differences in the voltage and current changes within each time period are analyzed, and the second characteristic value of each time period is determined; the second characteristic value is used to correct the soft threshold function in wavelet denoising. The modified wavelet denoising algorithm is used to denoise the electrical variables in each time period, and anomaly detection is performed on the denoised electrical variables to obtain the discharge test results of the COB light strip. The method of using the first eigenvalue to correct the soft threshold in each layer of wavelet denoising is expressed as follows: In the formula, The corrected soft threshold is used in the j-th layer wavelet denoising. Let be the first eigenvalue of the wavelet detail coefficients at the j-th level. This is a fixed threshold in wavelet denoising; The second characteristic value is the difference between the sum of the instantaneous phase changes of the voltage at all adjacent moments within each time period and the sum of the instantaneous phase changes of the current at all adjacent moments. The expression for modifying the soft threshold function in wavelet denoising using the second eigenvalue is as follows: In the formula, Let be the second characteristic value of the nth time interval, where Sig() is the Sigmoid function and sgn() is the sign function. These are the wavelet detail coefficients of the j-th level and k-th wavelet after wavelet decomposition. represents the k-th new wavelet detail coefficient of the j-th layer after soft thresholding.

2. The discharge test method for a COB lamp strip as described in claim 1, characterized in that, The determination of the noise impact weights for each maxima includes: Calculate the mean of the absolute values ​​of all wavelet detail coefficients remaining after removing the maxima within each local window of each layer. Calculate the difference between each maxima and the mean, and denote it as the first difference. The noise impact weight of each maxima is the proportion of the first difference of each maxima in the first difference of all maxima within its local window.

3. The discharge test method for a COB lamp strip as described in claim 1, characterized in that, The process of obtaining the noise impact weights for each local window includes: The noise impact weights of all maximum values ​​within each local window of each layer are sorted in descending order, and the fusion value of the first preset number of noise impact weights is used as the noise impact weight of each local window.

4. The discharge test method for a COB lamp strip as described in claim 1, characterized in that, The determination of the noise interference level of wavelet detail coefficients at each layer includes: By integrating the mean and dispersion of the noise impact weights of all local windows in each layer, the noise interference degree of the wavelet detail coefficients in each layer is obtained.

5. The discharge test method for a COB lamp strip as described in claim 1, characterized in that, The first eigenvalue is the sum of the normalized value of the energy entropy of all wavelet detail coefficients in each layer and the normalized value of the noise interference degree of the wavelet detail coefficients in each layer.

6. The discharge test method for a COB lamp strip as described in claim 1, characterized in that, The process of obtaining the discharge test results for the COB LED strip includes: If any outlier is found among all the electrical variables after noise reduction, the COB light strip is determined to have an electrical safety risk; otherwise, the COB light strip is determined not to have an electrical safety risk.

7. A discharge testing system for COB LED strips, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1-6.

Images