A method for monitoring health of photovoltaic components and supports of a mountain photovoltaic power station

Abstract

This invention relates to the field of health monitoring technology, specifically disclosing a method for health monitoring of photovoltaic modules and supports in mountainous photovoltaic power stations. The method involves connecting a capacitor and inductor in series between the positive and negative buses after disconnecting the inverter at night, injecting multiple sets of test voltages with superimposed dual frequencies and increasing amplitudes, and collecting the response signal on the negative terminal's ground sampling resistor. The amplitudes of the dual-frequency components and differential frequency intermodulation components are extracted through quadrature phase-locked loop (PLL). The ground parameters and differential frequency ground impedance modulus are solved by combining calibration coefficients and a parallel impedance model. Based on this, an equivalent nonlinear coefficient is calculated and fitted with the logarithm of the voltage amplitude. The nonlinear growth exponent is obtained by subtracting one from the slope to quantify the degree of micro-motion wear at the connection interface, achieving non-invasive online early warning.

Description

Technical Field

[0001] This invention relates to the field of health monitoring technology, specifically to a method for health monitoring of photovoltaic modules and supports in mountainous photovoltaic power stations. Background Technology

[0002] Mountain photovoltaic power stations are typically built on undulating, sloping terrain, with photovoltaic modules connected to the support structures using fasteners such as clamps and bolts. Under the long-term effects of wind-induced vibration, diurnal temperature variations, and snow accumulation, the interface between the module frame and the support structure undergoes reciprocating relative sliding at the micrometer level. This leads to wear of the surface oxide layer, a decrease in contact normal pressure, and the accumulation of wear debris—a process known as fretting wear. Fretting wear can cause safety hazards such as increased contact resistance, localized overheating, and even module detachment.

[0003] Traditional health monitoring methods mainly rely on manual inspections to observe the loosening of fasteners or the detection of abnormal heat points through infrared thermal imaging. However, these methods are all post-event detections and cannot detect the degradation trend of interface performance in the early stages of wear. In terms of electrical monitoring, existing technologies mainly measure the insulation resistance to ground of photovoltaic arrays, but there is no direct and definite mapping relationship between changes in insulation resistance and the degree of fretting wear. Therefore, there is an urgent need for a non-invasive online monitoring method that can quantitatively assess the fretting wear state of the interface between the module and the support during power plant operation breaks. Summary of the Invention

[0004] The purpose of this invention is to provide a method for health monitoring of photovoltaic modules and supports in mountain photovoltaic power stations, and to solve the following technical problems.

[0005] The objective of this invention can be achieved through the following technical solutions: A method for health monitoring of photovoltaic modules and supports in mountain photovoltaic power stations includes the following steps: Disconnect the photovoltaic module string from the DC side of the inverter at night, and connect a capacitor and an inductor in series between the positive bus and the negative bus; Prepare at least three sets of test voltages. Each set of test voltages consists of the superposition of sinusoidal voltages at the first and second frequencies, with the voltage amplitudes of different sets of test voltages increasing sequentially. Inject the test voltage into the series branch of the capacitor and inductor, connect a sampling resistor between the negative bus and ground, and collect the response voltage signal when the test voltage is injected from both ends of the sampling resistor. Extract the amplitude and phase of the first frequency component and the second frequency component, as well as the amplitude of the difference frequency intermodulation component, from the response voltage signal; The extracted amplitude and phase are converted into voltage components between the positive and negative buses. Combined with the test voltage and the series impedance of the capacitor and inductor, the magnitude of the impedance to ground at the difference between the first and second frequencies is calculated. The equivalent nonlinear coefficient is calculated based on the amplitude of the differential frequency intermodulation component and the modulus of the impedance to ground. The logarithmic coefficient and the logarithmic voltage value are obtained by taking the logarithm of the equivalent nonlinear coefficient and the voltage amplitude of the test voltage, respectively. A linear fit is then performed based on the logarithmic voltage value and the logarithmic coefficient value corresponding to all test voltages. Subtracting 1 from the slope of the straight line yields the nonlinear growth index. Based on the nonlinear growth index, the monitoring results of fretting wear at the connection interface between the photovoltaic module and the support are output.

[0006] As a further aspect of the present invention: during the connection of the capacitor and inductor: The monitoring system of the photovoltaic power station reads the current time. If the current time is within the time interval of 30 minutes after sunset to 30 minutes before sunrise, the monitoring system sends a disconnect command to the DC side electronic switch of the photovoltaic inverter. The DC side electronic switch is connected in series between the positive bus of the photovoltaic module string and the DC input terminal of the inverter. The monitoring system detects the voltage value between the positive bus and the negative bus through a voltage sensor. When the voltage value continues to drop to 5% below the open circuit voltage value of the photovoltaic module string, it confirms that the DC side electronic switch is completely disconnected. A monitoring branch is connected in parallel between the positive bus and the negative bus. The monitoring branch consists of a first relay, a capacitor, an inductor, and a second relay connected in series. After the DC side electronic switch is opened, the monitoring system sends a closing command to the first relay and the second relay at the same time, so that the capacitor and inductor are connected in series between the positive bus and the negative bus.

[0007] As a further aspect of the present invention: during the preparation of the test voltage: The first and second direct digital frequency synthesizers inside the monitoring system generate a first sine wave digital sequence and a second sine wave digital sequence, respectively. The frequencies of the first and second sine wave digital sequences are equal to the preset first frequency and second frequency, respectively. The absolute value of the difference between the second frequency and the first frequency is greater than 10 Hz and less than 100 Hz. The voltage amplitudes of the first and second sinusoidal digital sequences are generated according to at least three preset amplitude values. Each set of test voltages corresponds to one amplitude value, and the amplitude values ​​of different sets of test voltages increase sequentially. The first sine wave digital sequence and the second sine wave digital sequence under the same amplitude value are added point by point to obtain the superimposed digital sequence. The superimposed digital sequence is converted into an analog superimposed voltage by a digital-to-analog converter. The analog superimposed voltage is amplified by a power amplifier and output as a set of test voltages. The output impedance of the power amplifier is <1Ω.

[0008] As a further aspect of the present invention: during the process of acquiring the response voltage signal: The output of the test voltage is connected in parallel to both ends of the series branch of the capacitor and inductor; The first and second ends of the sampling resistor are connected to the negative busbar of the photovoltaic module string and the grounding busbar of the photovoltaic power station, respectively. The resistance value of the sampling resistor is 10Ω to 100Ω. A differential amplifier circuit is connected in parallel across the sampling resistor. The two input terminals of the differential amplifier circuit are connected to the first and second terminals of the sampling resistor, respectively. The output terminal of the differential amplifier circuit is connected to the input terminal of a bandpass filter. The output terminal of the bandpass filter is connected to the input channel of an analog-to-digital converter. The analog-to-digital converter outputs a digital sequence of response voltage signals. Among them, the lower passband frequency of the bandpass filter is lower than the difference between the first frequency and the second frequency, and the upper passband frequency is higher than the larger of the first frequency and the second frequency. The sampling frequency of the analog-to-digital converter is set to more than ten times the larger of the first frequency and the second frequency.

[0009] As a further aspect of the present invention: during the extraction of amplitude, phase, and difference frequency intermodulation component amplitude: Generate a reference sine sequence and a reference cosine sequence with a frequency equal to the first frequency. The phase zero of the reference sine sequence is aligned with the starting phase of the test voltage, and the reference cosine sequence lags behind the reference sine sequence by 1 / 4 cycle. The digital sequence of the response voltage signal is multiplied point by point with the reference sine sequence and the reference cosine sequence, and then summed to obtain the in-phase summation value and the quadrature summation value of the first frequency component. The amplitude and phase of the first frequency component are calculated from the in-phase summation value and the quadrature summation value. Obtain the amplitude and phase of the second frequency component; Generate a reference sine sequence and a reference cosine sequence whose frequencies are equal to the difference between the first frequency and the second frequency. Multiply the digital sequence of the response voltage signal with these two sequences point by point and sum them to obtain the in-phase accumulation value and the quadrature accumulation value of the difference frequency intermodulation component. Calculate the amplitude of the difference frequency intermodulation component from the in-phase accumulation value and the quadrature accumulation value.

[0010] As a further aspect of the present invention: in the process of calculating the modulus of impedance to ground: A calibration test shall be performed on the first night after the photovoltaic power station is put into operation for the first time, and after confirming that the connection interface between the photovoltaic modules and the support is in good condition. During the calibration test, the DC side electronic switch is disconnected and the first and second relays are closed. Single-frequency sinusoidal voltages with the first and second frequencies are injected into the series branch of the capacitor and inductor, respectively. The amplitude and phase of the response voltage across the sampling resistor are measured, and the ratio between the complex amplitude of the injected voltage and the complex amplitude of the response voltage at the sampling resistor is taken as the voltage conversion coefficient at that frequency.

[0011] As a further aspect of the present invention: in the process of calculating the modulus of impedance to ground: Complex numbers A1 and A2 are constructed based on the amplitude and phase of the first and second frequency components, respectively. The complex numbers A1 and A2 are multiplied by the corresponding voltage conversion coefficients to obtain the first and second complex voltage numbers of the positive and negative buses under the first and second frequency components. The ground network between the positive and negative busbars is equivalent to a capacitor and a resistor connected in parallel. The capacitance value of the capacitor is denoted as the equivalent capacitance to ground, and the resistance value of the resistor is denoted as the equivalent leakage resistance. Based on the first complex voltage, the complex amplitude of the first frequency component in the test voltage, and the complex value of the series impedance of the capacitor and inductor at the first frequency, the first complex equation is established. Based on the second complex voltage, the complex amplitude of the second frequency component in the test voltage, and the complex value of the series impedance of the capacitor and inductor at the second frequency, establish the second complex equation; Solving the simultaneous complex equations yields the equivalent capacitance to ground and the equivalent leakage resistance. Based on these values, the complex impedance of the network to ground at the difference between the first and second frequencies is calculated. The magnitude of this complex impedance is then taken as the magnitude of the impedance to ground at the difference between the first and second frequencies.

[0012] As a further aspect of the present invention: during the process of linear fitting: Based on the voltage conversion coefficients at the first and second frequencies, linear interpolation is used to calculate the voltage conversion coefficient B at the difference between the first and second frequencies. Divide the amplitude of the difference frequency intermodulation component by the magnitude of the voltage conversion coefficient B, and then divide it by the magnitude of the impedance to ground to obtain the equivalent nonlinear coefficient. Using base 10, take the logarithm of the voltage amplitude of the test voltage to obtain the logarithmic voltage value, and take the logarithm of the equivalent nonlinear coefficient corresponding to the set of test voltages to obtain the logarithmic coefficient value; At least three sets of test voltages yield at least three logarithmic voltage values ​​and at least three logarithmic coefficient values, and the at least three logarithmic voltage values ​​and the at least three logarithmic coefficient values ​​are all different. Using the logarithmic voltage value as the x-axis and the logarithmic coefficient value as the y-axis, at least three data points are obtained in a Cartesian coordinate system. The least squares method is then used to perform univariate linear regression on the at least three data points.

[0013] As a further aspect of the present invention: during the process of obtaining monitoring results: On the first night after the photovoltaic power plant is put into operation for the first time, the nonlinear growth index is obtained and used as the initial nonlinear growth index. If the difference between the current nonlinear growth index and the initial nonlinear growth index is greater than the preset growth threshold, an alarm message will be sent to the operation and maintenance personnel through the monitoring system of the photovoltaic power station. The alarm message contains the number of the photovoltaic module string that has experienced micro-dip wear and the value of the current nonlinear growth index.

[0014] The beneficial effects of this invention compared to the prior art are as follows: This invention utilizes the window of nighttime when photovoltaic modules do not generate electricity. After disconnecting the inverter on the DC side, a test voltage consisting of dual-frequency superposition is injected into the positive and negative buses. By collecting the response signal in the negative bus-to-ground circuit and extracting the differential frequency intermodulation component, quantitative detection of the nonlinear characteristics of the interface between the module and the support is achieved. Since fretting wear directly causes an enhancement of the current-voltage nonlinearity at the contact interface, the amplitude of the differential frequency intermodulation component and its growth law with the test voltage can accurately reflect the degree of wear, making the monitoring results unaffected by the drift of insulation parameters caused by changes in ambient temperature and humidity. This invention only requires the addition of a small number of passive devices and signal acquisition circuits to the existing photovoltaic electrical circuit, without the need to install sensors on each module, resulting in low construction costs and no impact on normal daytime power generation. By fitting logarithmic data under multiple sets of test voltages to obtain a nonlinear growth exponent, the degree of fretting wear can be transformed into a single comparable quantitative indicator, facilitating maintenance personnel to set early warning thresholds and realizing the transformation from periodic inspections to condition-oriented maintenance, effectively reducing the safety risks caused by support connection failures. Attached Figure Description

[0015] The invention will now be further described with reference to the accompanying drawings.

[0016] Figure 1 This is a flowchart illustrating a method for monitoring the health of photovoltaic modules and supports in a mountain photovoltaic power station according to the present invention. Figure 2 This is a schematic diagram of the process for calculating the impedance modulus to ground in this invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figures 1-2 As shown, this invention provides a method for health monitoring of photovoltaic modules and supports in mountainous photovoltaic power stations, comprising the following steps: Disconnect the photovoltaic module string from the DC side of the inverter at night, and connect a capacitor and an inductor in series between the positive bus and the negative bus; In a preferred embodiment of the present invention, during the connection of the capacitor and inductor: The monitoring equipment within the photovoltaic power station reads the current time and compares it with preset sunset and sunrise times. If the current time falls within the range of 30 minutes after sunset to 30 minutes before sunrise, the monitoring equipment sends a disconnect command to the DC-side electronic switch connected between the positive bus of the photovoltaic module string and the DC input terminal of the inverter. The DC-side electronic switch is a normally closed type; upon receiving the disconnect command, its internal power semiconductor device turns off, cutting off the electrical path for the photovoltaic module string to supply DC to the inverter.

[0019] The monitoring equipment continuously acquires voltage values ​​using a voltage sensor connected between the positive and negative busbars. After the DC-side electronic switch is opened, the voltage between the positive and negative busbars, initially close to the inverter's maximum power point tracking voltage, begins to decrease. The rate of decrease is affected by the discharge of the photovoltaic module's junction capacitance and ground capacitance. When the voltage value continuously decreases and eventually stabilizes below 5% of the open-circuit voltage of the photovoltaic module string, the monitoring equipment confirms that the DC-side electronic switch has been completely opened, and only residual voltage exists between the positive and negative busbars, with no continuous power output.

[0020] After confirmation of disconnection, the monitoring equipment simultaneously sends a closing command to both the first and second relays. The first and second relays, along with the capacitor and inductor connected in series between them, form a monitoring branch, which is connected in parallel between the positive and negative busbars. One end of the first relay is connected to the positive busbar, the other end is connected to the first terminal of the capacitor, the second terminal of the capacitor is connected to the first terminal of the inductor, the second terminal of the inductor is connected to one end of the second relay, and the other end of the second relay is connected to the negative busbar. Both the first and second relays are normally open relays; their contacts separate when no closing command is applied, disconnecting the electrical connection between the monitoring branch and the busbar. After the monitoring equipment simultaneously sends a closing command to both relays, the contacts of both relays simultaneously close, connecting the capacitor and inductor in series between the positive and negative busbars. For example, a film capacitor with a withstand voltage not less than 1.2 times the open-circuit voltage of the photovoltaic module series can be used, and an air-core inductor is used to avoid magnetic saturation due to DC bias.

[0021] Prepare at least three sets of test voltages. Each set of test voltages consists of the superposition of sinusoidal voltages at the first and second frequencies, with the voltage amplitudes of different sets of test voltages increasing sequentially. In a preferred embodiment of the present invention, during the preparation of the test voltage: The monitoring equipment is equipped with a first direct digital frequency synthesizer that generates a first sine wave digital sequence based on a preset first frequency, for example, 1020 Hz. A second direct digital frequency synthesizer generates a second sine wave digital sequence based on a preset second frequency, for example, 1080 Hz. The absolute value of the difference between the second and first frequencies is greater than 10 Hz and less than 100 Hz. Each of the first and second direct digital frequency synthesizers quantizes each sample point of the sine wave into a 16-bit signed integer and temporarily stores it in a buffer memory.

[0022] The voltage amplitudes of the first and second sinusoidal digital sequences are generated with the same amplitude value. The monitoring equipment pre-stores at least three distinct amplitude values, such as 3V, 5V, and 8V. For the first amplitude value of 3V, a first direct digital frequency synthesizer generates a first sinusoidal digital sequence with a peak value of 3V, and a second direct digital frequency synthesizer generates a second sinusoidal digital sequence with a peak value of 3V. A digital adder adds the first and second sinusoidal digital sequences with the same amplitude value, sample by sample, to obtain a superimposed digital sequence.

[0023] The superimposed digital sequence is fed into the input register of the digital-to-analog converter (DAC). The DAC converts each data word in the superimposed digital sequence into a corresponding analog voltage step value based on the sampling clock. After smoothing by a reconstruction low-pass filter at the output, a continuous analog superimposed voltage is formed. This analog superimposed voltage contains two sinusoidal components with frequencies of 1020 Hz and 1080 Hz.

[0024] The simulated superimposed voltage is connected to the input of a power amplifier configured as a voltage follower topology, whose open-loop output impedance is reduced to less than 1Ω after deep negative feedback. The output of the power amplifier serves as the output port for a set of test voltages. After sequentially outputting a test voltage with an amplitude of 3 volts, the monitoring device reads the next amplitude value of 5 volts, and the above process is repeated to generate the second set of test voltages, until all test voltage groups corresponding to the preset amplitude values ​​have been sequentially generated and output.

[0025] Inject the test voltage into the series branch of the capacitor and inductor, connect a sampling resistor between the negative bus and ground, and collect the response voltage signal when the test voltage is injected from both ends of the sampling resistor. In another preferred embodiment of the present invention, during the acquisition of the response voltage signal: The output of the test voltage is directly connected in parallel to the two ends of the series branch of the capacitor and inductor via a wire. When the capacitor and inductor are connected between the positive and negative busbars through a relay, the test voltage is applied to this series branch.

[0026] The first end of the sampling resistor is soldered to the negative busbar copper bus of the photovoltaic module string, and the second end is connected to the grounding busbar of the photovoltaic power station through a grounding wire. The resistance value of the sampling resistor is selected as 50Ω, which is in the range of 10Ω to 100Ω.

[0027] A differential amplifier circuit is connected in parallel across the sampling resistor. The first input terminal of the differential amplifier circuit is connected to the first terminal of the sampling resistor, and the second input terminal is connected to the second terminal of the sampling resistor. The differential amplifier circuit adopts an instrumentation amplifier structure with a common-mode rejection ratio greater than 100 dB. It is used to extract the weak differential-mode voltage signal across the sampling resistor while suppressing common-mode interference.

[0028] The output of the differential amplifier circuit is connected to the input of the bandpass filter via a coaxial cable. The lower passband cutoff frequency of the bandpass filter is set to 40 Hz, which is 60 Hz lower than the difference between the first frequency (1020 Hz) and the second frequency (1080 Hz). The upper passband cutoff frequency of the bandpass filter is set to 2000 Hz, which is higher than the larger of the first and second frequencies, 1080 Hz. The bandpass filter is used to attenuate 50 Hz power frequency interference and high-frequency switching noise.

[0029] The output of the bandpass filter is connected to the analog input channel of the analog-to-digital converter (ADC). The ADC's sampling frequency is set to 12000 Hz, which is more than ten times the larger of the first and second frequencies, 1080 Hz. Driven by the sampling clock, the ADC converts the filtered analog voltage into a 16-bit integer digital value and outputs a digital sequence of the response voltage signal through a serial peripheral interface. This digital sequence is temporarily stored in a dual-port random access memory (DRAM) for subsequent signal processing units to read and extract amplitude and phase.

[0030] Extract the amplitude and phase of the first frequency component and the second frequency component, as well as the amplitude of the difference frequency intermodulation component, from the response voltage signal; In a preferred embodiment of the present invention, during the extraction of amplitude, phase, and difference frequency intermodulation component amplitude: The digital signal processor internally generates a reference sine sequence and a reference cosine sequence with a frequency equal to the first frequency of 1020 Hz. The phase zero point of the reference sine sequence is strictly aligned with the start time of the test voltage output. Each sample value of the reference cosine sequence lags behind the corresponding sample value of the reference sine sequence by a quarter cycle.

[0031] The digital sequence of the response voltage signal is read into the multiply-accumulate unit of the digital signal processor (DSP). Each sample point of the digital sequence is multiplied point-by-point by the corresponding sample point of the reference sine sequence, and the sum of all products is obtained to get the in-phase accumulated value of the first frequency component. Each sample point of the digital sequence is then multiplied point-by-point by the corresponding sample point of the reference cosine sequence, and the sum is obtained to get the quadrature accumulated value of the first frequency component. The DSP then calculates the sum of the squares of the in-phase accumulated value and the squares of the quadrature accumulated value, and takes the square root of this sum. The result is the amplitude of the first frequency component. Simultaneously, the ratio of the quadrature accumulated value to the in-phase accumulated value is calculated, and the arctangent of this ratio is taken; the resulting radian value is the phase of the first frequency component.

[0032] The same processing procedure is used to obtain the amplitude and phase of the second frequency component, which will not be elaborated on here.

[0033] A digital signal processor generates a reference sequence with a frequency equal to the difference between a first frequency and a second frequency, which is 60 Hz. The phase zeros of the reference sine and cosine sequences are aligned with the starting phase of the test voltage. The digital sequence of the response voltage signal is multiplied point-by-point by the 60 Hz reference sine and cosine sequences and accumulated to obtain the in-phase and quadrature accumulated values ​​of the difference frequency intermodulation component. The square root of the sum of the squares of the in-phase and quadrature accumulated values ​​of the difference frequency intermodulation component is calculated to obtain the amplitude of the difference frequency intermodulation component.

[0034] Understandably, the reference sine and cosine sequences generated internally by the digital signal processor are equivalent to two local oscillation signals with a 90-degree phase difference. When the response voltage signal is multiplied point-by-point with the reference sequences and then accumulated, this mathematical operation is equivalent to calculating the Fourier coefficients of the response signal at a specific frequency. For any interference component with a frequency different from that of the reference sequence, the result of multiplication and accumulation will approach zero due to the periodic cancellation of positive and negative values. Only signal components with frequencies strictly consistent with the reference sequence can produce a non-zero accumulated value.

[0035] The in-phase accumulation value reflects the projected length of the frequency component in the zero-phase reference direction, while the quadrature accumulation value reflects the projected length of the frequency component in the quadrature direction. Summing the squares of both and taking the square root yields the true amplitude of the frequency component, and the arctangent of their ratio gives the offset angle of the frequency component relative to the reference zero phase. Performing this operation on both the first and second frequencies allows for the precise separation of the response amplitude and phase of each injected frequency from a response signal containing various noises.

[0036] The difference frequency component is also extracted using quadrature phase-locked loop (PLL), and its frequency is equal to the difference between the first and second frequencies. When there is no nonlinear distortion at the connection interface between the photovoltaic module and the support, the superposition and injection of two sinusoidal voltages of different frequencies will only generate a current response in the circuit with the same injection frequency, without generating new frequency components. However, when the connection interface experiences contact loosening and oxide layer accumulation due to fretting wear, the current and voltage no longer maintain a linear proportional relationship. At this time, the dual-frequency voltage excitation will generate intermodulation products through the nonlinear characteristics of the interface, among which the most significant and easily detectable component is the difference frequency component. Therefore, the magnitude of the difference frequency intermodulation component directly indicates the strength of the nonlinearity of the connection interface. By extracting this magnitude and processing it together with the subsequently calculated ground impedance modulus and voltage conversion coefficient, the nonlinear growth index characterizing the fretting wear state can be derived.

[0037] The extracted amplitude and phase are converted into voltage components between the positive and negative buses. Combined with the test voltage and the series impedance of the capacitor and inductor, the magnitude of the impedance to ground at the difference between the first and second frequencies is calculated. In a preferred embodiment of this invention, during the calculation of the ground impedance modulus: On the first night after the photovoltaic power station is put into operation, it is confirmed that the connection interface between the photovoltaic modules and the support is in good condition, that is, the tightening torque of all clamping bolts meets the design value and infrared thermal imaging shows no abnormal hot spots. At this time, a calibration test is performed. After disconnecting the DC side electronic switch and closing the first and second relays, the capacitor C and the inductor L are connected in series between the positive and negative busbars.

[0038] First, a single-frequency sinusoidal voltage with a first frequency of f1 is injected into the series branch of the capacitor and inductor. The complex amplitude of this injected voltage is denoted as Uin1, which is the peak value of the injected voltage, and the phase is referenced to zero at the injection moment. Simultaneously, the response voltage across the sampling resistor Rs is measured, and the resistance value of the sampling resistor is denoted as Rs. After differential amplification, bandpass filtering, and analog-to-digital conversion, the response voltage is processed by quadrature phase-locked loop (PLL) to obtain its complex amplitude, denoted as Urs1. The voltage conversion coefficient K1 = Uin1 / Urs1 at the first frequency f1 is calculated.

[0039] Next, a single-frequency sinusoidal voltage with a second frequency f2 is injected, and the complex amplitude of the injected voltage is denoted as Uin2. Similarly, the complex amplitude of the response voltage Urs2 across the sampling resistor is measured. The voltage conversion coefficient K2 = Uin2 / Urs2 at the second frequency f2 is calculated.

[0040] During normal monitoring, a test voltage consisting of the superposition of a first frequency f1 and a second frequency f2 is injected into the series branch of the capacitor and inductor. The test voltage is obtained by generating a first sine wave digital sequence and a second sine wave digital sequence through a direct digital frequency synthesizer inside the monitoring equipment, superimposing them, and then performing digital-to-analog conversion and power amplification.

[0041] The voltage amplitude of the first frequency component is preset to Um1, and the initial phase is preset to θ1. Based on this, the complex amplitude of the first frequency component in the test voltage is constructed. j is the imaginary unit, and the square of j equals -1.

[0042] The voltage amplitude of the second frequency component is preset to Um2, and the initial phase is preset to θ2. Based on this, the complex amplitude of the second frequency component in the test voltage is constructed. .

[0043] The output of the test voltage is directly connected in parallel across the two ends of the series branch of capacitor C and inductor L, and also in parallel between the positive and negative busbars. Therefore, the voltage between the positive and negative busbars is equal to the test voltage itself. At the first frequency f1, the complex bus voltage V1 is equal to Ut1, and at the second frequency f2, the complex bus voltage V2 is equal to Ut2.

[0044] The response voltage signal acquired from both ends of the sampling resistor Rs is differentially amplified, bandpass filtered, and converted from analog to digital. Then, through quadrature phase-locked loop (PLL) processing, the complex amplitude A1 of the first frequency component and the complex amplitude A2 of the second frequency component are extracted. Both A1 and A2 are complex numbers, containing the voltage amplitude and phase information generated by the frequency component across the sampling resistor.

[0045] in, M1 and P1 represent the amplitude and phase of the first frequency component, respectively. M2 and P2 represent the amplitude and phase of the second frequency component, respectively.

[0046] The sampling resistor Rs is connected in series between the negative busbar of the photovoltaic module string and the grounding busbar of the photovoltaic power station. Therefore, the current flowing through the sampling resistor Rs is exactly equal to the current flowing through the equivalent network to ground. According to Ohm's law, the complex impedance Zg1 of the network to ground at the first frequency f1 can be directly calculated by the following formula: Zg1 = Rs * (Ut1 / A1) = Rs * K1; similarly, the complex impedance Zg2 of the network to ground at the second frequency f2 = Rs * (Ut2 / A2) = Rs * K2.

[0047] In the above two equations, Ut1 and Ut2 are known complex amplitudes of the injection voltage, A1 and A2 are measured complex amplitudes of the response voltage, and Rs is a known resistance value of the sampling resistor. Therefore, Zg1 and Zg2 are both known complex numbers that can be directly calculated through measurement.

[0048] The ground network between the positive and negative busbars is equivalent to a model of a capacitor and a resistor connected in parallel. The equivalent capacitance to ground is denoted as CO, and the equivalent leakage resistance is denoted as R0. At frequency f1, the complex impedance of the ground network is denoted as Zg1, and its expression is: ; At frequency f2, the complex impedance of the network to ground is denoted as Zg2, and its expression is: ; Substituting Zg1 and Zg2, obtained through measurement calculations, into the two impedance expressions above, we obtain two independent complex equations containing unknowns R0 and C0. Each complex equation can be decomposed into two real equations with real and imaginary parts. However, since the unknowns are only two real numbers, R0 and C0, the system of equations is exact and has a unique solution. For example, we can take the reciprocals of both equations to obtain the admittance form, separate the real and imaginary parts, and then solve the system of equations simultaneously.

[0049] Specifically, for the first frequency f1: ; For the first frequency f2, we have: ; Taking the reciprocals of the known Zg1 and Zg2, their real parts are equal to the reciprocals of R0, and their imaginary parts are equal to 2πf1C0 and 2πf2CO, respectively. From this, the reciprocal of the equivalent leakage resistance R0 can be directly calculated, thus yielding R0. Dividing the imaginary part value at any frequency by 2πf1 or 2πf2 gives the equivalent capacitance to ground CO. For example, CO is obtained by dividing the imaginary part value corresponding to f1 by 2πf1. Thus, the two characteristic parameters R0 and C0 of the ground network are accurately solved.

[0050] After obtaining C0 and R0, calculate the frequency at the difference between the first and second frequencies, denoted as fd, where fd equals the absolute value of 12 minus 11. At frequency fd, the complex impedance of the network to ground is denoted as Zgd, and its expression is: ; The magnitude Zd of Zgd characterizes the impedance of the photovoltaic module string to ground network under differential frequency fd excitation. Since R0 and CO are accurate parameters obtained based on the actual circuit topology, Zd can truly reflect the ground impedance characteristics at this frequency point. This value will be used in subsequent steps to normalize the equivalent nonlinear coefficients, so as to eliminate the influence of differences in the ground insulation conditions of different photovoltaic module strings on the nonlinearity detection results.

[0051] It should be noted that the sampling resistor is connected in series between the negative bus and the grounding bus. The current flowing through the sampling resistor is uniquely equal to the total current discharged from the positive and negative buses to the ground through the equivalent ground network, excluding the current component flowing through the series branch of the capacitor and inductor. When the test voltage is applied directly between the positive and negative buses, the bus voltage is the known test voltage. Dividing the response voltage across the sampling resistor by the resistance value yields the instantaneous value of the ground current. Then, the complex impedance of the ground network at the corresponding frequency can be directly calculated from the complex ratio of the test voltage to the ground current. Electrically, the ground network can be equivalently represented as a parallel structure of a capacitor and a resistor. This equivalent model can accurately describe the leakage and capacitive coupling path from the photovoltaic module frame through the support to the ground in the low- to mid-frequency range. After measuring the complex impedance to ground at two different frequencies, these values ​​are substituted into the mathematical expression for parallel impedance. Since the expression only contains two unknown real parameters—equivalent leakage resistance and equivalent capacitance to ground—the two independent complex equations provide sufficient constraints to solve for these two unknowns. The solution process utilizes the separation of the real and imaginary parts of the reciprocal of the complex number. The real part corresponds to the reciprocal of the equivalent leakage resistance, while the imaginary part is proportional to the product of the frequency and the equivalent capacitance to ground. This allows for the precise calculation of the equivalent leakage resistance and equivalent capacitance to ground. Once these two parameters are obtained, the impedance to ground at any frequency can be directly calculated using the parallel model, including the magnitude of the impedance to ground at the difference frequency between the two injection frequencies. This magnitude reflects the magnitude of the impedance to ground of the photovoltaic module string under the current insulation state at a specific difference frequency, and its value is unaffected by subsequent changes in the injection voltage amplitude. Dividing the amplitude of the differential frequency intermodulation component by the ground impedance modulus is equivalent to eliminating the difference in response sensitivity caused by the difference in the insulation characteristics of the photovoltaic module itself to the ground. This allows the final calculated equivalent nonlinear coefficient to simply reflect the degree of nonlinear distortion caused by fretting wear at the connection interface between the photovoltaic module and the support.

[0052] The equivalent nonlinear coefficient is calculated based on the amplitude of the differential frequency intermodulation component and the modulus of the impedance to ground. The logarithmic coefficient and the logarithmic voltage value are obtained by taking the logarithm of the equivalent nonlinear coefficient and the voltage amplitude of the test voltage, respectively. A linear fit is then performed based on the logarithmic voltage value and the logarithmic coefficient value corresponding to all test voltages. In a preferred embodiment of the present invention, during the linear fitting process: Based on the voltage conversion coefficient K1 at the first frequency f1 and the voltage conversion coefficient K2 at the second frequency f2, linear interpolation is used to calculate the voltage conversion coefficient B at the difference between the first and second frequencies. The interpolation method uses frequency as the independent variable, treating K1 and K2 as complex values ​​at frequencies f1 and f2. Interpolation is performed proportionally to frequency in the complex plane to obtain the real and imaginary parts of the voltage conversion coefficient B at frequency fd.

[0053] Dividing the previously extracted difference-frequency intermodulation component amplitude A3 by the magnitude of the voltage conversion coefficient B, and then dividing the quotient by the magnitude of the impedance to ground Zd, yields the equivalent nonlinear coefficient, denoted as D. This operation eliminates the contribution of the measurement loop gain and the linear impedance to ground to the difference-frequency signal amplitude.

[0054] At least three preset test voltages, each with a different voltage amplitude, are used, for example, 3 volts, 5 volts, and 8 volts. The aforementioned process is repeated for each test voltage group, yielding equivalent nonlinear coefficients D1, D2, and D3. The logarithmic voltage values, denoted as X1, X2, and X3, are obtained by taking the logarithm of the voltage amplitude Um of each test voltage group, with base 10. Simultaneously, the corresponding equivalent nonlinear coefficients D1, D2, and D3 are also taken as logarithmic coefficient values, denoted as Y1, Y2, and Y3. For example, when Um is 3 volts, X equals lg3, and when D has a certain value, Y equals lgD.

[0055] Using each set of logarithmic voltage values ​​as the x-axis and the corresponding logarithmic coefficient values ​​as the y-axis, at least three data points are formed in a Cartesian coordinate system. A univariate linear regression is performed on these data points using the least squares method, and the slope and intercept of the regression line are calculated. The fitting process obtains the slope value that minimizes the sum of squared residuals by solving the normal equations.

[0056] It is important to note that obtaining the voltage conversion coefficient at the difference frequency is to convert the measured amplitude of the difference frequency intermodulation component back to the true nonlinear response voltage at both ends of the positive and negative bus. Since the difference frequency itself is not a pre-injected frequency, there is no directly corresponding calibration data available. However, the voltage conversion coefficients at the first and second frequencies are known. The two frequencies are numerically located on opposite sides of the difference frequency. Therefore, linear interpolation can be used to approximate the transfer characteristics at the intermediate frequency using the frequency as a bridge, thereby avoiding distortion of the nonlinear amplitude due to the gain of the measurement channel.

[0057] The difference frequency amplitude is then divided by the magnitude of the voltage conversion coefficient and then by the magnitude of the impedance to ground. This division essentially removes linear factors. The difference frequency signal originates from the nonlinear distortion at the connection interface, but as it flows through the entire loop, it is subjected to both amplitude-frequency response of the measurement channel and linear impedance to ground. The magnitude of the voltage conversion coefficient reflects the attenuation ratio of the signal from the bus end to the sampling end, while the magnitude of the impedance to ground determines the voltage drop across the sampling resistor under the same nonlinear current. After successively eliminating these two items, the resulting equivalent nonlinear coefficient is only related to the nonlinearity of the connection interface itself, and is decoupled from linear interferences such as the drift of the leakage resistance to ground caused by changes in ambient humidity and the capacitance fluctuation caused by temperature.

[0058] Repeating the above measurements with at least three sets of test voltages of different amplitudes yields equivalent nonlinear coefficients for multiple excitation intensities. The base-10 logarithmic operation stems from the power-law characteristic of the contact interface nonlinearity. Theory shows that the amplitude of the intermodulation product and the excitation voltage amplitude are approximately linearly related in a double logarithmic coordinate system, and the slope directly relates to the order characteristics of the nonlinearity. Plotting points with the logarithmic voltage as the x-axis and the logarithmic coefficient as the y-axis and fitting a straight line using the least squares method effectively extracts the statistical growth trend of nonlinearity with increasing excitation from a finite number of discrete excitation points, rather than relying on the accidental values ​​of single-point measurements. Subtracting one from the slope of the fitted straight line is to remove the baseline contribution of the linear component from the order of the intermodulation components, resulting in a normalized dimensionless index for the nonlinear growth. A larger index indicates a greater deviation of the current-voltage relationship at the connection interface from Ohm's law, meaning more severe contact loosening and oxidation caused by fretting wear. This transforms complex electrical measurements into a single numerical value, facilitating maintenance personnel to set thresholds and issue warnings accordingly.

[0059] Subtracting 1 from the slope of the straight line yields the nonlinear growth index. Based on this index, the monitoring results of fretting wear at the interface between the photovoltaic module and the support are output. In a preferred embodiment of the present invention, during the process of obtaining monitoring results: On the first night after the photovoltaic power plant is put into operation for the first time, the nonlinear growth index is obtained and used as the initial nonlinear growth index. If the difference between the current nonlinear growth index and the initial nonlinear growth index is greater than the preset growth threshold, an alarm message will be sent to the operation and maintenance personnel through the monitoring system of the photovoltaic power station. The alarm message contains the number of the photovoltaic module string that has experienced micro-dip wear and the value of the current nonlinear growth index.

[0060] It is important to note that in this invention, after disconnecting the inverter at night, an injection branch consisting of a capacitor and an inductor is connected in series between the positive and negative buses of the photovoltaic module. A test voltage with dual-frequency superposition and increasing amplitude is used to excite the photovoltaic array's ground loop. Due to fretting wear, a nonlinear current-voltage characteristic is generated at the interface between the module and the support. The dual-frequency excitation generates a difference frequency intermodulation component through this nonlinear interface, the magnitude of which is directly related to the degree of interface wear. The difference frequency component is accurately extracted by acquiring the response signal on the negative terminal ground sampling resistor and using quadrature phase-locked loop technology. Simultaneously, the single-frequency injection measured during the calibration phase... The voltage conversion coefficient is combined with the parallel impedance model to solve for the equivalent capacitance to ground and leakage resistance, thereby obtaining the impedance modulus to ground at the difference frequency. The amplitude of the difference frequency component is divided by this impedance modulus to eliminate the influence of insulation differences between different arrays and obtain the equivalent nonlinear coefficient. Then, the nonlinear coefficients obtained under multiple sets of test voltages with different amplitudes are logarithmically fitted to a straight line. The slope of the straight line minus one is used as the nonlinear growth index. This index can quantitatively characterize the degree of nonlinear distortion caused by fretting wear at the connection interface, thereby realizing non-invasive online health monitoring and early warning without installing sensors on each component.

[0061] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. A method for health monitoring of photovoltaic modules and supports in mountain photovoltaic power stations, characterized in that, Includes the following steps: Disconnect the photovoltaic module string from the DC side of the inverter at night, and connect a capacitor and an inductor in series between the positive bus and the negative bus; Prepare at least three sets of test voltages. Each set of test voltages consists of the superposition of sinusoidal voltages at the first and second frequencies, with the voltage amplitudes of different sets of test voltages increasing sequentially. Inject the test voltage into the series branch of the capacitor and inductor, connect a sampling resistor between the negative bus and ground, and collect the response voltage signal when the test voltage is injected from both ends of the sampling resistor. Extract the amplitude and phase of the first frequency component and the second frequency component, as well as the amplitude of the difference frequency intermodulation component, from the response voltage signal; The extracted amplitude and phase are converted into voltage components between the positive and negative buses. Combined with the test voltage and the series impedance of the capacitor and inductor, the magnitude of the impedance to ground at the difference between the first and second frequencies is calculated. The equivalent nonlinear coefficient is calculated based on the amplitude of the differential frequency intermodulation component and the modulus of the impedance to ground. The logarithmic coefficient and the logarithmic voltage value are obtained by taking the logarithm of the equivalent nonlinear coefficient and the voltage amplitude of the test voltage, respectively. A linear fit is then performed based on the logarithmic voltage value and the logarithmic coefficient value corresponding to all test voltages. Subtracting 1 from the slope of the straight line yields the nonlinear growth index. Based on the nonlinear growth index, the monitoring results of fretting wear at the connection interface between the photovoltaic module and the support are output.

2. The method for health monitoring of photovoltaic modules and supports in a mountain photovoltaic power station according to claim 1, characterized in that, During the process of connecting capacitors and inductors: The monitoring system of the photovoltaic power station reads the current time. If the current time is within the time interval of 30 minutes after sunset to 30 minutes before sunrise, the monitoring system sends a disconnect command to the DC side electronic switch of the photovoltaic inverter. The DC side electronic switch is connected in series between the positive bus of the photovoltaic module string and the DC input terminal of the inverter. The monitoring system detects the voltage value between the positive bus and the negative bus through a voltage sensor. When the voltage value continues to drop to 5% below the open circuit voltage value of the photovoltaic module string, it confirms that the DC side electronic switch is completely disconnected. A monitoring branch is connected in parallel between the positive bus and the negative bus. The monitoring branch consists of a first relay, a capacitor, an inductor, and a second relay connected in series. After the DC side electronic switch is opened, the monitoring system sends a closing command to the first relay and the second relay at the same time, so that the capacitor and inductor are connected in series between the positive bus and the negative bus.

3. The method for health monitoring of photovoltaic modules and supports in a mountain photovoltaic power station according to claim 1, characterized in that, During the preparation of the test voltage: The first and second direct digital frequency synthesizers inside the monitoring system generate a first sine wave digital sequence and a second sine wave digital sequence, respectively. The frequencies of the first and second sine wave digital sequences are equal to the preset first frequency and second frequency, respectively. The absolute value of the difference between the second frequency and the first frequency is greater than 10 Hz and less than 100 Hz. The voltage amplitudes of the first and second sinusoidal digital sequences are generated according to at least three preset amplitude values. Each set of test voltages corresponds to one amplitude value, and the amplitude values ​​of different sets of test voltages increase sequentially. The first sine wave digital sequence and the second sine wave digital sequence under the same amplitude value are added point by point to obtain the superimposed digital sequence. The superimposed digital sequence is converted into an analog superimposed voltage by a digital-to-analog converter. The analog superimposed voltage is amplified by a power amplifier and output as a set of test voltages. The output impedance of the power amplifier is <1Ω.

4. The method for health monitoring of photovoltaic modules and supports in a mountain photovoltaic power station according to claim 1, characterized in that, During the process of acquiring the response voltage signal: The output of the test voltage is connected in parallel to both ends of the series branch of the capacitor and inductor; The first and second ends of the sampling resistor are connected to the negative busbar of the photovoltaic module string and the grounding busbar of the photovoltaic power station, respectively. The resistance value of the sampling resistor is 10Ω to 100Ω. A differential amplifier circuit is connected in parallel across the sampling resistor. The two input terminals of the differential amplifier circuit are connected to the first and second terminals of the sampling resistor, respectively. The output terminal of the differential amplifier circuit is connected to the input terminal of a bandpass filter. The output terminal of the bandpass filter is connected to the input channel of an analog-to-digital converter. The analog-to-digital converter outputs a digital sequence of response voltage signals. Among them, the lower passband frequency of the bandpass filter is lower than the difference between the first frequency and the second frequency, and the upper passband frequency is higher than the larger of the first frequency and the second frequency. The sampling frequency of the analog-to-digital converter is set to more than ten times the larger of the first frequency and the second frequency.

5. A method for health monitoring of photovoltaic modules and supports in a mountain photovoltaic power station according to claim 1, characterized in that, During the extraction of amplitude, phase, and difference frequency intermodulation component amplitudes: Generate a reference sine sequence and a reference cosine sequence with a frequency equal to the first frequency. The phase zero of the reference sine sequence is aligned with the starting phase of the test voltage, and the reference cosine sequence lags behind the reference sine sequence by 1 / 4 cycle. The digital sequence of the response voltage signal is multiplied point by point with the reference sine sequence and the reference cosine sequence, and then summed to obtain the in-phase summation value and the quadrature summation value of the first frequency component. The amplitude and phase of the first frequency component are calculated from the in-phase summation value and the quadrature summation value. Obtain the amplitude and phase of the second frequency component; Generate a reference sine sequence and a reference cosine sequence whose frequencies are equal to the difference between the first frequency and the second frequency. Multiply the digital sequence of the response voltage signal with these two sequences point by point and sum them to obtain the in-phase accumulation value and the quadrature accumulation value of the difference frequency intermodulation component. Calculate the amplitude of the difference frequency intermodulation component from the in-phase accumulation value and the quadrature accumulation value.

6. The method for health monitoring of photovoltaic modules and supports in a mountain photovoltaic power station according to claim 1, characterized in that, In the process of calculating the magnitude of the impedance to ground: A calibration test shall be performed on the first night after the photovoltaic power station is put into operation for the first time, and after confirming that the connection interface between the photovoltaic modules and the support is in good condition. During the calibration test, the DC side electronic switch is disconnected and the first and second relays are closed. Single-frequency sinusoidal voltages with the first and second frequencies are injected into the series branch of the capacitor and inductor, respectively. The amplitude and phase of the response voltage across the sampling resistor are measured, and the ratio between the complex amplitude of the injected voltage and the complex amplitude of the response voltage at the sampling resistor is taken as the voltage conversion coefficient at that frequency.

7. A method for health monitoring of photovoltaic modules and supports in a mountain photovoltaic power station according to claim 6, characterized in that, In the process of calculating the magnitude of the impedance to ground: Complex numbers A1 and A2 are constructed based on the amplitude and phase of the first and second frequency components, respectively. The complex numbers A1 and A2 are multiplied by the corresponding voltage conversion coefficients to obtain the first and second complex voltage numbers of the positive and negative buses under the first and second frequency components. The ground network between the positive and negative busbars is equivalent to a capacitor and a resistor connected in parallel. The capacitance value of the capacitor is denoted as the equivalent capacitance to ground, and the resistance value of the resistor is denoted as the equivalent leakage resistance. Based on the first complex voltage, the complex amplitude of the first frequency component in the test voltage, and the complex value of the series impedance of the capacitor and inductor at the first frequency, the first complex equation is established. Based on the second complex voltage, the complex amplitude of the second frequency component in the test voltage, and the complex value of the series impedance of the capacitor and inductor at the second frequency, establish the second complex equation; Solving the simultaneous complex equations yields the equivalent capacitance to ground and the equivalent leakage resistance. Based on these values, the complex impedance of the network to ground at the difference between the first and second frequencies is calculated. The magnitude of this complex impedance is then taken as the magnitude of the impedance to ground at the difference between the first and second frequencies.

8. A method for health monitoring of photovoltaic modules and supports in a mountain photovoltaic power station according to claim 1, characterized in that, During the process of linear fitting: Based on the voltage conversion coefficients at the first and second frequencies, linear interpolation is used to calculate the voltage conversion coefficient B at the difference between the first and second frequencies. Divide the amplitude of the difference frequency intermodulation component by the magnitude of the voltage conversion coefficient B, and then divide it by the magnitude of the impedance to ground to obtain the equivalent nonlinear coefficient. Using base 10, take the logarithm of the voltage amplitude of the test voltage to obtain the logarithmic voltage value, and take the logarithm of the equivalent nonlinear coefficient corresponding to the set of test voltages to obtain the logarithmic coefficient value; At least three sets of test voltages yield at least three logarithmic voltage values ​​and at least three logarithmic coefficient values, and the at least three logarithmic voltage values ​​and the at least three logarithmic coefficient values ​​are all different. Using the logarithmic voltage value as the x-axis and the logarithmic coefficient value as the y-axis, at least three data points are obtained in a Cartesian coordinate system. The least squares method is then used to perform univariate linear regression on the at least three data points.

9. A method for health monitoring of photovoltaic modules and supports in a mountain photovoltaic power station according to claim 1, characterized in that, During the process of obtaining monitoring results: On the first night after the photovoltaic power plant is put into operation for the first time, the nonlinear growth index is obtained and used as the initial nonlinear growth index. If the difference between the current nonlinear growth index and the initial nonlinear growth index is greater than the preset growth threshold, an alarm message will be sent to the operation and maintenance personnel through the monitoring system of the photovoltaic power station. The alarm message contains the number of the photovoltaic module string that has experienced micro-dip wear and the value of the current nonlinear growth index.

Images