A Fault Detection Method for Power Inverter Based on Output Signal Analysis

Abstract

The present invention provides a power inverter fault detection method based on output signal analysis, wherein the power inverter fault detection method collects power inverter input and output signals, first makes a preliminary judgment based on the input and output electric energy difference, then uniformly samples the output voltage signal within a period at equal intervals, performs wavelet decomposition and reconstruction, extracts feature quantities, and then judges whether a fault occurs and the specific fault location. The present invention provides a power inverter fault detection method based on output signal analysis, which can accurately judge whether a power inverter fails and improve work reliability.

Description

Technical Field

[0001] The invention belongs to the technical field of power detection, and in particular relates to a power inverter fault detection method based on output signal analysis. Background Art

[0002] With the advancement of science and technology and the improvement of industrial production levels, people have increasingly stringent requirements on the reliability and safety of power equipment. Some equipment needs to ensure its power supply. If the power is disconnected, it may cause the safety of equipment and personnel.

[0003] In order to ensure that some important equipment is powered on continuously, a DC power supply is required so that the important equipment can be powered in time when the system is powered off. As an important component of the safe power supply system, the power inverter converts the DC power into AC power and provides the AC power that meets the requirements for important equipment in time. Therefore, it is necessary to ensure that the power inverter works reliably and accurately.

[0004] The present invention proposes a power inverter fault detection method based on output signal analysis, which collects the input and output signals of the power inverter, first makes a preliminary judgment based on the input and output electric energy difference, then uniformly samples the output voltage signal within a cycle at equal intervals, extracts the characteristic quantity, and then determines whether a fault occurs and the specific fault location. Summary of the invention

[0005] The present invention provides a power inverter fault detection method based on output signal analysis, which can accurately determine whether the power inverter has a fault and improve working reliability.

[0006] The present invention is specifically a power inverter fault detection method based on output signal analysis, and the power inverter fault detection method comprises the following steps:

[0007] Step (1): collecting the input voltage signal, input current signal, output voltage signal, and output current signal of the power inverter;

[0008] Step (2): Calculating the input power and output power of the power inverter;

[0009] Step (3): Calculating the difference between the input electric energy and the output electric energy;

[0010] Step (4): determine whether the difference is greater than a difference reference value, if so, proceed to step (5), if not, return to step (1);

[0011] Step (5): extracting the voltage signal of the entire cycle through a phase-locked loop frequency multiplication circuit;

[0012] Step (6): performing three-layer wavelet decomposition on the voltage signal db3 wavelet basis to obtain a low-frequency coefficient and three high-frequency coefficients;

[0013] Step (7): reconstruct the wavelet decomposition coefficients of each node to obtain the reconstructed signal;

[0014] Step (8): Calculate the energy sum of each decomposition coefficient;

[0015] Step (9): constructing a feature vector according to each of the decomposition coefficients;

[0016] Step (10): normalize the feature vector to obtain a normalized feature vector:

[0017] Step (11): Determine whether each component of the normalized eigenvector is greater than an energy reference value. If so, the power inverter is faulty and proceed to step (12); if not, return to step (5);

[0018] Step (12): selecting the minimum value of each component of the normalized eigenvector;

[0019] Step (13): Determine the specific fault location according to the minimum value of each component of the normalized eigenvector.

[0020] The energy sum of the decomposition coefficients is x i is the reconstructed signal amplitude; the eigenvector is

[0021] use The feature vector is normalized.

[0022] The phase-locked loop frequency multiplication circuit adopts a method of uniformly sampling the voltage signal at equal intervals within a cycle, which can achieve synchronous sampling of the entire cycle, avoid spectrum leakage in the signal acquisition link, and improve the accuracy of signal acquisition.

[0023] Compared with the prior art, the beneficial effect is: the power inverter fault detection method collects the input and output signals of the power inverter, first makes a preliminary judgment based on the input and output electric energy difference, then uniformly samples the output voltage signal within a cycle at equal intervals, extracts the characteristic quantity, and then determines whether a fault occurs and the specific fault location. DETAILED DESCRIPTION

[0024] The specific implementation of a power inverter fault detection method based on output signal analysis of the present invention is described in detail below.

[0025] The power inverter fault detection method of the present invention comprises the following steps: (1): collecting the input voltage signal, input current signal, output voltage signal and output current signal of the power inverter; (2): calculating the input electric energy and output electric energy of the power inverter; (3): calculating the difference between the input electric energy and the output electric energy; (4): judging whether the difference is greater than the difference reference value, if so, entering (5), if not, returning to (1); (5): extracting the whole cycle voltage signal through the phase-locked loop frequency multiplication circuit; (6): performing three-layer wavelet decomposition on the voltage signal db3 wavelet basis to obtain a low-frequency coefficient and three high-frequency coefficients; (7): reconstructing the wavelet decomposition coefficients of each node to obtain a reconstructed signal; (8): calculating the energy and x i To reconstruct the signal amplitude; (9): Construct the feature vector according to each decomposition coefficient (10): Normalize the feature vector to obtain the normalized feature vector (11): Determine whether each component of the normalized eigenvector is greater than the energy reference value. If so, the power inverter is faulty and go to (12); if not, return to (5); (12): Select the minimum value of each component of the normalized eigenvector; (13): Determine the specific fault location based on the minimum value of each component of the normalized eigenvector.

[0026] The phase-locked loop frequency multiplication circuit adopts the method of uniformly sampling the voltage signal at equal intervals within a cycle, which can realize synchronous sampling of the entire cycle, avoid spectrum leakage in the signal acquisition link, and improve the accuracy of signal acquisition.

[0027] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention and are not intended to limit the present invention. A person skilled in the art should understand that the specific implementation of the present invention can be modified or replaced by equivalents, but these modifications or changes are within the scope of protection of the pending claims.

Claims

1. A power inverter fault detection method based on output signal analysis, characterized in that: The power inverter fault detection method comprises the following steps: Step (1): collecting the input voltage signal, input current signal, output voltage signal, and output current signal of the power inverter; Step (2): Calculating the input power and output power of the power inverter; Step (3): Calculating the difference between the input electric energy and the output electric energy; Step (4): determine whether the difference is greater than a difference reference value, if so, proceed to step (5), if not, return to step (1); Step (5): extracting the voltage signal of the entire cycle through a phase-locked loop frequency multiplication circuit; Step (6): performing three-layer wavelet decomposition on the voltage signal db3 wavelet basis to obtain a low-frequency coefficient and three high-frequency coefficients; Step (7): reconstruct the wavelet decomposition coefficients of each node to obtain the reconstructed signal; Step (8): Calculate the energy sum of each decomposition coefficient; the energy sum of each decomposition coefficient is x i is the reconstructed signal amplitude; Step (9): construct a feature vector according to each decomposition coefficient; the feature vector is Step (10): normalize the feature vector to obtain a normalized feature vector: The feature vector is normalized. Step (11): Determine whether each component of the normalized eigenvector is greater than an energy reference value. If so, the power inverter is faulty and proceed to step (12); if not, return to step (5); Step (12): selecting the minimum value of each component of the normalized eigenvector; Step (13): Determine the specific fault location according to the minimum value of each component of the normalized eigenvector.

2. A power inverter fault detection method based on output signal analysis according to claim 1, characterized in that: The phase-locked loop frequency multiplication circuit adopts a method of uniformly sampling the voltage signal at equal intervals within a cycle, which can achieve synchronous sampling of the entire cycle.

Images