Photovoltaic power station remote monitoring system based on Internet of Things

Abstract

The invention relates to the technical field of solar photovoltaic power generation and electric energy conversion control, in particular to a photovoltaic power station remote monitoring system based on the Internet of Things, and the system comprises a data collection module which is used for monitoring the illumination intensity, temperature and output power of a photovoltaic module; the data analysis module is used for determining the output power at the next moment of the current moment, and comprises the following steps: determining an illumination influence factor and a temperature influence factor; correcting the predicted value of the output power at the next moment according to the predicted values of the illumination intensity and the temperature at the next moment, the illumination influence factor and the temperature influence factor; and the monitoring module is used for accurately and timely monitoring the photovoltaic power station according to the difference between the corrected predicted value of the output power at the next moment and the actual value.

Description

Technical Field

[0001] The present invention relates to the technical field of solar photovoltaic power generation and power conversion control. Specifically, it relates to a remote monitoring system for photovoltaic power stations based on the Internet of Things. Background Art

[0002] As a clean and renewable energy technology, photovoltaic power stations have become an important part of the global energy transformation. They convert infrared, visible, and ultraviolet light in solar radiation into electrical energy through photovoltaic modules, and then use inverters to convert direct current into alternating current for use by households and enterprises or to be integrated into the power grid. With the diversification of electricity demand and the increasing requirements for grid stability, monitoring and controlling the output power of photovoltaic power stations is of great significance for improving power conversion efficiency and ensuring grid stability.

[0003] In the prior art, time series prediction algorithms are usually used to predict future data based on historical data, and then determine whether the data is abnormal according to the difference between the future predicted value and the actual value, thereby realizing the monitoring function. For example, the ARIMA (Autoregressive Integrated Moving Average) prediction algorithm is a commonly used time series prediction algorithm.

[0004] However, the power output of a photovoltaic power station is not only affected by the performance of the equipment, but also strongly interfered by environmental factors, which will affect the power generation efficiency of the photovoltaic modules. When using traditional time series prediction algorithms to monitor photovoltaic power stations, only historical output power data is relied on for linear fitting to predict the output power at future moments. This prediction algorithm fails to consider the influence of environmental factors on the output power, and the error of the prediction results of this prediction algorithm will gradually increase, especially in the case of drastic changes in light or sudden weather changes, resulting in a significant deviation between the predicted value of the output power of the photovoltaic power station and the actual value of the output power, and it is impossible to accurately and timely monitor the photovoltaic power station. Summary of the Invention

[0005] To solve the problem that when monitoring the output power of a photovoltaic power station, the existing method uses traditional time series prediction algorithms to predict only based on historical output power data and fails to consider the influence of environmental factors on the output power, resulting in the inability to accurately and timely monitor the photovoltaic power station, the present invention proposes a remote monitoring system for photovoltaic power stations based on the Internet of Things, which includes: A data acquisition module for monitoring the light intensity, temperature, and output power of photovoltaic modules; A data analysis module for determining the output power at the next moment of the current moment, including: Predict the light intensity, temperature and output power at the next moment through a prediction algorithm based on the historical light intensity, historical temperature and historical output power at each moment; obtain multiple high-frequency components and low-frequency components of the historical light intensity, historical temperature and historical power through empirical mode decomposition technology; Obtain the influence factors of each high-frequency component of the historical light intensity and historical temperature on each high-frequency component of the historical output power, and the influence factors of each low-frequency component of the historical light intensity and historical temperature on each low-frequency component of the historical output power, and determine the light influence factor and temperature influence factor through weighted average operation; Correct the predicted value of the output power at the next moment according to the predicted values of the light intensity and temperature at the next moment, as well as the light influence factor and temperature influence factor; A monitoring module for remotely monitoring the operating state of the photovoltaic power station according to the difference between the corrected predicted value and the actual value of the output power at the next moment.

[0006] This technical solution first predicts the light intensity, temperature and output power at the next moment of the current moment through a traditional prediction algorithm, and then through empirical mode decomposition technology, it can finely capture the data characteristics at different time scales, avoiding prediction errors caused by complex environmental fluctuations or emergencies. By calculating the influence factors of the high-frequency components and low-frequency components of environmental factors on the high-frequency components and low-frequency components of the output power, it fully reflects the influence of light and temperature changes on the output power at different time scales. Different environmental factors will produce different effects at different time periods, and the system can identify and quantify this influence. And through weighted average operation, the influence factors of light and temperature are comprehensively evaluated, which helps to balance the influence of different factors on the output power of the photovoltaic power station and optimize the prediction effect of the model. And on the basis of the original prediction result, by introducing the correction of environmental factors, the accuracy of the predicted result of the output power is improved, enabling the prediction algorithm to better cope with environmental changes, avoiding prediction errors caused by temperature or light fluctuations, making the prediction result more accurate, and realizing accurate remote monitoring of the photovoltaic power station according to the difference between the accurate prediction result and the actual situation.

[0007] Furthermore, the influence factor of each high-frequency component of the historical light intensity and historical temperature on each high-frequency component of the historical output power is determined based on the following method: Select any high-frequency component of the historical light intensity or historical temperature as the first high-frequency component, and any high-frequency component of the historical output power as the second high-frequency component; Calculate all the change rates of the first high-frequency component and all the change rates of the second high-frequency component respectively according to the numerical values of the first high-frequency component and the second high-frequency component at all historical moments; Calculate the covariance of all the change rates of the first high-frequency component and all the change rates of the first high-frequency component, as well as the variance of all the change rates of the first high-frequency component, and use the ratio of the covariance to the variance as the influence factor of the first high-frequency component on the second high-frequency component.

[0008] This technical solution can reveal the rapidly changing components in the data through the analysis of high-frequency components. These components have a direct impact on the fluctuations of the output power. The rapid changes in temperature or light intensity within a short period will directly affect the rapid fluctuations of the output power. Through the differential and covariance analysis of high-frequency fluctuations, the change relationship between the sudden changes in the environment and the output power of the photovoltaic power station can be accurately captured.

[0009] Furthermore, the influence factor of each low-frequency component of the historical light intensity and historical temperature on each low-frequency component of the historical output power is determined based on the following method: Select any low-frequency component of the historical light intensity or historical temperature as the first low-frequency component, and any low-frequency component of the historical output power as the second low-frequency component; Take the sum of the values of the first low-frequency component and the second low-frequency component at each historical moment and all previous moments as the cumulative values of the first low-frequency component and the second low-frequency component at that historical moment; Calculate the covariance of the cumulative values of the first low-frequency component at all historical moments and the cumulative values of the second low-frequency component at all historical moments, as well as the variance of the cumulative values of the first low-frequency component at all historical moments; use the ratio of the covariance to the variance as the influence factor of the first low-frequency component on the second low-frequency component.

[0010] This technical solution can quantify the long-term influence of historical light intensity and historical temperature on historical output power by analyzing the cumulative values of low-frequency components and calculating the ratio of covariance to variance as the influence factor. This method can fully reflect the continuous influence of the long-term change trend of the environment on the output power, making the prediction result of the output power more accurate.

[0011] Furthermore, the method for determining the light influence factor and temperature influence factor through weighted average operation includes: Perform weighted summation according to the weights of each high-frequency component of the historical light intensity / historical temperature and the influence factors of each high-frequency component of the historical light intensity / historical temperature on each high-frequency component of the historical power to obtain the first light cumulative value / first temperature cumulative value; Perform weighted summation according to the weights of each low-frequency component of the historical light intensity / historical temperature and the influence factors of each low-frequency component of the historical light intensity / historical temperature on each low-frequency component of the historical power to obtain the second light cumulative value / second temperature cumulative value; Take the sum of the first light accumulation value / first temperature accumulation value and the second light accumulation value / second temperature accumulation value as the light influence factor / temperature influence factor.

[0012] Through the weighted average operation, this technical solution can dynamically adjust the predicted value according to the specific changes in light and temperature. When the environment changes drastically, the weight of the high-frequency component increases, so that the monitoring system pays more attention to these sudden changes and avoids the deviation of the predicted value. When the environment changes relatively smoothly, the influence of the low-frequency component will be amplified, so as to improve the prediction accuracy of the long-term trend, ensure that the monitoring system can flexibly respond to different types of environmental fluctuations, and improve the reliability of long-term and short-term predictions.

[0013] Furthermore, the weight is determined based on the following method: The weight of each high-frequency component / low-frequency component of the historical light intensity is the variance of the values of this high-frequency component / low-frequency component at all historical moments, divided by the sum of the variances of the values of all high-frequency components and all low-frequency components of the historical light intensity at all historical moments; The weight of each high-frequency component / low-frequency component of the historical temperature is the variance of the values of this high-frequency component / low-frequency component at all historical moments, divided by the sum of the variances of the values of all high-frequency components and all low-frequency components of the historical temperature at all historical moments.

[0014] Furthermore, a method for correcting the predicted value of the output power at the next moment is based on the following formula: ; In the formula, is the predicted value of the power corrected at the next moment of the current moment, is the predicted value of the output power at the next moment of the current moment, is the light influence factor, is the temperature influence factor, is the predicted value of the light intensity at the next moment of the current moment, is the predicted value of the temperature at the next moment of the current moment, is the mean value of the historical light intensity at the current moment, is the mean value of the historical temperature at the current moment.

[0015] By introducing the influence factors of light and temperature, two environmental factors, on power output, this technical solution can make the prediction result of the output power more in line with the actual situation. Under high light intensity, it has a positive influence on the output power, while in a high-temperature environment, it has an inhibitory effect on the output power. By dynamically adjusting the predicted value through the influence of environmental factors, the prediction result is more in line with the actual situation of the photovoltaic power station.

[0016] Further, another method for correcting the predicted value of the output power at the next moment is based on the following formula: ; In the formula, is the predicted value of the power correction at the next moment of the current moment, is the predicted value of the output power at the next moment of the current moment, is the light influence factor, is the temperature influence factor, is the predicted value of the light intensity at the next moment of the current moment, is the predicted value of the temperature at the next moment of the current moment, is the average value of the historical light intensity at the current moment, is the average value of the historical temperature at the current moment, is the light saturation effect correction term, is the temperature attenuation correction term.

[0017] This technical solution further improves the correction process of the power prediction of the photovoltaic power station. The addition of the light saturation effect correction term and the temperature attenuation correction term is to handle the limiting effects of light and temperature on power output. When the light intensity is too high, the light saturation effect correction term can prevent the model from predicting too high a power output when the light is too strong. When the temperature is too high, the temperature attenuation correction term is used to accelerate the power attenuation to conform to the physical characteristics of the photovoltaic modules, resulting in a more accurate power prediction result.

[0018] Further, the light saturation effect correction term is calculated based on the following formula: ; In the formula, is the light saturation effect correction term, is the light intensity saturation value, is the predicted value of the light intensity at the next moment of the current moment, is the natural exponential function.

[0019] Further, the temperature attenuation correction term is determined based on the following formula: ; In the formula, is the temperature attenuation correction term, is the predicted value of the light intensity at the next moment of the current moment, is the temperature critical value, is the natural exponential function.

[0020] Further, the method for obtaining the temperature critical value is: Pre - test the output power of the photovoltaic module at different temperatures; construct a coordinate system with the abscissa being temperature and the ordinate being output power, and obtain the curve of the output power varying with temperature; calculate the slope at each point on the fitted curve, and take the temperature value corresponding to the minimum negative slope as the temperature critical value.

[0021] The present invention has the following effects: By precisely capturing the environmental characteristics at different time scales, introducing the influence of high - frequency and low - frequency components of environmental factors on the power output of the photovoltaic power station, comprehensively considering the complexity and dynamics of environmental changes, the present invention improves the accuracy of output power prediction, enabling the monitoring system to accurately and effectively monitor the photovoltaic power station based on accurate prediction results, and ensuring the stable operation of the photovoltaic power station. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 is the system structure block diagram of the present invention.

[0023] Figure 2 is the flow schematic diagram for predicting the output power of the photovoltaic power station of the present invention; Figure 3 is the flow schematic diagram for analyzing and calculating the light influence factor and the temperature influence factor of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0024] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention.

[0025] A remote monitoring system for a photovoltaic power station based on the Internet of Things provided by the present invention, as Figure 1 shown in, includes: a data acquisition module S1, a data analysis module S2, and a monitoring module S3.

[0026] Data acquisition module S1: This module includes Internet of Things devices, specifically temperature sensors, light intensity sensors installed on photovoltaic modules, and photovoltaic inverters. Photovoltaic inverters in photovoltaic power stations usually have real - time monitoring and data acquisition functions, and can monitor key electrical parameters such as the output power, current, and voltage of photovoltaic modules. At each moment, a light intensity, a temperature, and an output power are collected.

[0027] The data such as the output power, light intensity, and temperature of the monitored photovoltaic power station are transmitted to the data analysis module S2 in real - time through a wireless network. The data analysis module S2 can accurately predict the output power through data analysis.

[0028] The output power of the photovoltaic modules in a photovoltaic power station usually has an approximately linear relationship with the light intensity, which is based on the principle of the photovoltaic effect. When the light intensity increases, the photon energy excites more electron-hole pairs, increasing the short-circuit current of the photovoltaic modules. At the same time, within a certain range, the open-circuit voltage also rises. Since the output power of the photovoltaic modules is closely related to the product of the current and voltage, the increase in light intensity directly leads to a significant increase in the output power.

[0029] The influence of temperature on power output: Different from the positive influence of light intensity, temperature has a negative influence on the power output of photovoltaic modules. Photovoltaic modules are usually composed of semiconductor materials (such as silicon), and their working principle depends on the movement of charge carriers (electrons and holes) in the electric field. When the temperature rises, the mobility of charge carriers in the semiconductor material is affected by thermal excitation, resulting in a decrease in its effective conductivity, thereby reducing the current output. In addition, the increase in temperature may also cause the bandgap of the photovoltaic module to narrow, further reducing the power output.

[0030] Data analysis module S2: This module includes a memory and a processor. A computer program is stored on the memory, and the processor executes the computer program to implement the following steps to predict the output power of the photovoltaic power station, as Figure 2 shown in, including steps S21 - S23: S21: Use the time series prediction algorithm to make a preliminary prediction of the output power of the photovoltaic power station at the next moment of the current moment.

[0031] During the operation of the photovoltaic power station, taking any moment as the current moment, all the moments before the current moment are regarded as all the historical moments of the current moment.

[0032] Obtain all the temperatures, light intensities, and output powers of the photovoltaic power station at all historical moments of the current moment as the historical temperature, historical light intensity, and historical power of the current moment. It can be seen that the historical temperature, historical light intensity, and historical power are all data sequences, including multiple values, and each value contained in the historical temperature, historical light intensity, and historical power corresponds one by one based on the acquisition moment.

[0033] Use the traditional ARIMA algorithm to obtain the predicted values of the temperature, light intensity, and output power of the photovoltaic power station at the next moment of the current moment according to the historical temperature, historical light intensity, and historical output power of the current moment.

[0034] S22: Analyze and calculate the light influence factor and the temperature influence factor.

[0035] The output power of a photovoltaic power station not only depends on the performance of the photovoltaic modules themselves but is also strongly affected by environmental factors. Light intensity and temperature are the most direct environmental factors, which have a significant impact on power output. Since these environmental factors may exhibit different fluctuation patterns on different time scales, simply relying on the linear fitting of historical data by traditional prediction algorithms is not sufficient to accurately reflect the actual operating conditions. Therefore, in this step, by analyzing the light influence factor and the temperature influence factor, the specific impact of environmental changes on the power output of the photovoltaic power station can be captured more precisely.

[0036] Specifically, as Figure 3 shown in, it includes steps S221 - S224: S221: Perform EMD decomposition on the historical light intensity, historical temperature, and historical output power at the current moment to obtain their respective high-frequency components and low-frequency components.

[0037] EMD (Empirical Mode Decomposition) is an adaptive time series data decomposition method that can decompose complex time series data into multiple components with different frequencies. Each component represents the fluctuations of the data on different time scales. Through EMD decomposition, the changes in the output power of the photovoltaic power station under different environmental conditions can be better captured. Especially in the case where it is greatly affected by environmental factors, it can effectively distinguish short-term changes from long-term trends. This process can help analyze the fluctuation patterns of power data more precisely and improve the accuracy of power prediction.

[0038] EMD decomposition will obtain multiple IMF (Intrinsic Mode Function) components. Each IMF component represents the fluctuations of the power data on different time scales. One IMF component is equivalent to a function curve with oscillatory characteristics. Subsequently, the IMF components will be simply referred to as components. These components consist of multiple high-frequency components and low-frequency components.

[0039] Usually, the results of EMD decomposition will arrange the components in the order from high frequency to low frequency. This is determined by the decomposition process of EMD. Through continuous screening, it gradually extracts the high-frequency components in the original signal to form high-frequency components. As the decomposition progresses, the frequencies of the subsequent obtained components gradually decrease, gradually forming low-frequency components.

[0040] To more accurately divide the high-frequency components and low-frequency components, it is set here that: If the number of components obtained after decomposition is odd, take the first components as high-frequency components and the remaining components as low-frequency components; if the number of components obtained after decomposition is even, take the first as high-frequency components and the remaining components as low-frequency components, where is the total number of data participating in the decomposition.

[0041] Therefore, the historical light intensity, historical temperature, and historical output power of the PV power station at the current moment are decomposed into multiple high-frequency components and multiple low-frequency components of the historical light intensity, multiple high-frequency components and multiple low-frequency components of the historical temperature, and multiple high-frequency components and multiple low-frequency components of the historical output power.

[0042] The historical light intensity at the current moment is denoted as , and is decomposed into high-frequency components , where represents the serial number, , and at the same time, low-frequency components are obtained: , where represents the serial number, .

[0043] The historical temperature at the current moment is denoted as , and is decomposed into high-frequency components , where represents the serial number, , and at the same time, low-frequency components are obtained: , where represents the serial number, .

[0044] The historical power at the current moment is denoted as , and is decomposed into high-frequency components , where represents the serial number, , and at the same time, low-frequency components are obtained: , where represents the serial number, .

[0045] For the decomposition results of the historical light intensity: The high-frequency components represent the rapid fluctuations of the light intensity, which are usually related to sudden changes in sunlight, weather phenomena (such as cloud cover), etc. The high-frequency components can capture the instantaneous changes in light, which have a direct impact on the power output of the PV power station. The low-frequency components represent the long-term change trend of the light intensity, which are usually related to long-term factors such as seasonal changes and sunshine duration. For example, the difference in sunlight intensity between summer and winter, the annual change in sunlight, etc.

[0046] For the decomposition results of historical temperature, the high-frequency component represents the short-term fluctuations of temperature data, such as weather changes, sudden high or low temperature phenomena, etc. The high-frequency component can capture the immediate impact of these rapidly fluctuating temperature changes on the power of the photovoltaic power station. The low-frequency component represents the long-term change trend of temperature data, which is usually related to seasonal changes or long-term climate changes in temperature. For example, the temperature difference between summer and winter, the annual temperature trend, etc.

[0047] For the decomposition results of historical power: The high-frequency component represents the rapidly changing part of the output power of the photovoltaic power station, which is usually closely related to factors such as short-term changes in light intensity and sudden weather events. For example, the instantaneous impact of factors such as cloud cover and wind speed changes on the output power of the photovoltaic power station can be captured by the high-frequency component. The low-frequency component represents the slower fluctuating part of the output power of the photovoltaic power station, which is usually related to seasonal changes in temperature or long-term environmental factors (such as climate change). The low-frequency component reflects the trend changes under the long-term stable operation of the photovoltaic power station.

[0048] In summary, the historical output power, historical temperature, and historical light intensity of the photovoltaic power station at the current moment are respectively decomposed into multiple high-frequency components and low-frequency components. This process can effectively distinguish short-term fluctuations from long-term trends and provide more accurate data support for subsequent power prediction. By capturing fluctuations on different time scales, the prediction model can better cope with environmental changes.

[0049] S222: Calculate the change rate of each high-frequency component and each low-frequency component at each moment.

[0050] In time series analysis, the change rate is an important indicator used to measure the speed and amplitude of data change over time. The output power of a photovoltaic power station is affected by environmental factors such as light and temperature, and these effects usually have different time scales. By calculating the change rate, the fluctuation characteristics of data such as light and temperature on short-term (high-frequency) and long-term (low-frequency) can be clearly captured.

[0051] In one embodiment, the calculation formula for the change rate of each high-frequency component of historical light intensity at each historical moment is:

[0052] In this formula, is the change rate of the th high-frequency component of historical light intensity at the th historical moment, is the value of the th high-frequency component of historical light intensity at moment (light intensity value), is the th high-frequency component of historical light intensity at the The numerical value of the moment (light intensity numerical value).

[0053] In one embodiment, the calculation formula for the change rate of each low-frequency component of the historical light intensity at each historical moment is:

[0054] In this formula, is the change rate of the th low-frequency component of the historical light intensity at the th historical moment, is the numerical value of the th low-frequency component of the historical light intensity at the th historical moment, is the numerical value of the th low-frequency component of the historical light intensity at the th historical moment.

[0055] In one embodiment, the change rate of the th high-frequency component of the historical temperature at the th historical moment is:

[0056] In this formula, is the change rate of the th high-frequency component of the historical temperature at the th historical moment, is the numerical value (temperature numerical value) of the th high-frequency component of the historical temperature at the th historical moment, is the numerical value (temperature numerical value) of the th high-frequency component of the historical temperature at the th historical moment.

[0057] In one embodiment, the change rate of each low-frequency component of the historical temperature at each historical moment is:

[0058] In this formula, is the change rate of the th low-frequency component of the historical temperature at the th historical moment, is the numerical value of the th high-frequency component of the historical temperature at the th historical moment, is the numerical value of the th low-frequency component of the historical temperature at the th historical moment.

[0059] In one embodiment, the change rate of each high-frequency component of the historical power at each historical moment is:

[0060] In this formula, is the change rate of the th high-frequency component of the historical power at the th historical moment, is the value (power value) of the th high-frequency component of the historical power at historical moment, is the value (power value) of the th high-frequency component of the historical power at the th historical moment.

[0061] In one embodiment, the change rate of each low-frequency component of the historical power at each historical moment is:

[0062] In this formula, is the change rate of the th low-frequency component of the historical power at the th moment, is the value of the th low-frequency component of the historical power at moment, is the value of the th low-frequency component of the historical power at the th moment.

[0063] In summary, the change rate of the high-frequency component can reflect the immediate impact of environmental factors such as instantaneous weather changes (such as cloud cover changes) on power output, while the change rate of the low-frequency component reflects the longer-term trend changes, such as the cumulative effect of seasonal changes on power output. By calculating these change rates, the fluctuation amplitudes of factors such as light and temperature can be quantified, which helps to accurately analyze their impact on power.

[0064] S223: Determine the influence factors between high-frequency components and the influence factors between low-frequency components.

[0065] Since the high-frequency components reflect fast-changing and instantaneously responsive dynamic behaviors, with strong randomness and instantaneousness, the correlation between high-frequency components can often be assumed by simple linearity. Therefore, by analyzing the co-variation trend between the change rates of the high-frequency components of light and temperature and the change rates of the high-frequency components of power, the short-term impact of light and temperature on the output power in the short term can be reflected.

[0066] Since the low-frequency components are more related to long-term trends and cumulative effects. The changes in low-frequency components are usually slow and long-term, and they may be affected by non-linear factors. For example, long-term changes in light intensity may cause changes in power output, and these changes will not be immediately reflected in the short term. Therefore, by analyzing the cumulative values of low-frequency components at each moment, the cumulative effects of light and temperature on the output power over a longer time scale can be obtained.

[0067] Therefore, in this step, by determining the influence factors of each high-frequency component of historical light intensity and historical temperature on each high-frequency component of historical power, and the influence factors of each low-frequency component of historical light intensity and historical temperature on each low-frequency component of historical power, the different effects of light and temperature on the output power of the photovoltaic power station at different time scales can be reflected. This analysis process can not only help identify the rapid effects of short-term environmental fluctuations on power output, but also reveal the continuous effects of long-term environmental changes on power output, providing an important basis for further power prediction correction.

[0068] In one embodiment, the influence factor of each high-frequency component of historical light intensity and historical temperature on each high-frequency component of historical output power is determined based on the following method: Select any high-frequency component of historical light intensity or historical temperature as the first high-frequency component, and any high-frequency component of historical output power as the second high-frequency component; Calculate all the change rates of the first high-frequency component and all the change rates of the second high-frequency component respectively according to the numerical values of the first high-frequency component and the second high-frequency component at all historical moments; Calculate the covariance of all the change rates of the first high-frequency component and all the change rates of the first high-frequency component, and the variance of all the change rates of the first high-frequency component, and take the ratio of the covariance and the variance as the influence factor of the first high-frequency component on the second high-frequency component.

[0069] Specifically, the calculation method of the influence factor of each high-frequency component of historical light intensity on each high-frequency component of historical power is:

[0070] In this formula, is the influence factor of the th high-frequency component of historical light intensity on the th high-frequency component of historical power, is the total number of historical moments at the current moment, is the change rate of the th high-frequency component of historical power at the th historical moment, is the th high-frequency component of historical light intensity at the The rate of change at a historical moment, The historical power The average of the rate of change of the high-frequency components at all historical moments, is the historical light intensity The mean of the rate of change of the high-frequency components at all historical moments.

[0071] If the impact factor is greater than 0, it indicates that the historical light intensity The high frequency component and the historical power The more positive correlation characteristics the high-frequency components show, and the larger the value of the influencing factor, the stronger the positive correlation characteristics, which is consistent with the physical characteristics of photovoltaic modules. For example, the movement of clouds causes a sudden increase in light intensity, and the power increases accordingly. When the light weakens, the output power also weakens.

[0072] If the impact factor is less than 0, it means that the historical light intensity The high frequency component and the historical power The more negative correlation characteristics the high-frequency components show, and the smaller the impact factor (negative number), the stronger the negative correlation characteristics, indicating that the increase in light intensity reduces the output power, indicating that the photovoltaic module may be faulty or blocked: such as inverter failure, dust accumulation on the solar panel. If the impact factor is equal to 0, it means that there is no obvious linear relationship between the change in light intensity and the change in power, and the change in light intensity is not obvious enough to affect the change in output power.

[0073] The numerator of this formula is the historical light intensity. The rate of change of the high-frequency component at all historical moments and the historical power The covariance of the change rate of the high-frequency component of the historical illumination intensity at all historical moments reflects the synchronous change trend between the change rate of the high-frequency component of the historical illumination intensity and the change rate of the high-frequency component of the historical power through the covariance. Specifically, if at a certain historical moment, the change rate of the high-frequency component of the historical illumination intensity and the change rate of the high-frequency component of the historical power increase or decrease at the same time, it means that the change trends of the two are consistent. The more such historical moments, the larger the covariance will be, indicating that the synchronous change trend is better and the impact factor will be larger. If at a certain historical moment, the change rate of the high-frequency component of the historical illumination intensity and the change rate of the high-frequency component of the historical power show one increase and the other decrease, it means that the change trends of the two are opposite. The more such historical moments, the smaller the covariance will be, indicating that the synchronous change trend is worse and the impact factor will be smaller.

[0074] The denominator of this formula is the historical light intensity. The variance of the change rate of the high-frequency component at all historical moments measures the fluctuation amplitude of the high-frequency component of the historical light intensity. The larger the variance, the larger the fluctuation amplitude. When the fluctuation of the high-frequency component of the historical light intensity is very small, that is, the variance in the denominator is small, it means that the amplitude of the light change is relatively small or relatively stable. Usually, when calculating the covariance and variance, it is necessary to divide by the total number of samples, that is, the numerator part of this formula needs to be divided by , and the denominator part also needs to be divided by , so the in the numerator and denominator parts cancel each other out.

[0075] Overall, when the light change is relatively stable (the denominator is smaller), if the synchronous trend between the power change and the light change is stronger (the numerator part is larger), it means that any small change in light will have a greater impact on the power change, and the influence factor value will be larger. On the contrary, if the light changes violently (the denominator is larger), then even if the synchronous effect between the light and the power is very good (the numerator is larger), this synchronous effect will be diluted. Because the violent fluctuation of the light makes the influence of the power fluctuation less prominent, and the value of the influence factor is relatively small.

[0076] To sum up, the influence factor calculated by this formula reveals the relationship strength between the light intensity and the power, reflects the actual influence of the light intensity change on the output power of the photovoltaic power station, and further provides a theoretical basis for power prediction and system optimization.

[0077] Specifically, the calculation method of the influence factor of each high-frequency component of the historical temperature on each high-frequency component of the historical power is as follows:

[0078] In this formula, is the influence factor of the th high-frequency component of the historical temperature on the th high-frequency component of the historical power, is the total number of historical moments at the current moment, is the change rate of the th high-frequency component of the historical power at the th historical moment, is the change rate of the th high-frequency component of the historical temperature at the th historical moment, is the average value of the change rates of the th high-frequency component of the historical power at all historical moments, is the average value of the change rates of the th high-frequency component of the historical temperature at all historical moments.

[0079] If the impact factor calculated by this formula is greater than 0, it indicates that the th high-frequency component of the historical temperature and the th high-frequency component of the historical power show a more positive correlation. Moreover, the larger the value of the impact factor, the stronger the positive correlation, which is less in line with the physical characteristics of the photovoltaic module. Because the power of the photovoltaic module is suppressed as the temperature rises. At this time, it indicates that there may be a fault in the system, such as a temperature sensor failure, resulting in incorrect measurement data, or there is a problem with the inverter, or part of the photovoltaic module is blocked or the surface is contaminated, resulting in abnormal changes in power output, and even an incorrect response opposite to the temperature change.

[0080] If the impact factor is less than 0, it means that the th high-frequency component of the historical temperature and the th high-frequency component of the historical power show a more negative correlation. And the smaller the impact factor (negative number), the stronger the negative correlation, indicating that the increase in temperature reduces the output power, which is in line with the physical characteristics of the photovoltaic module.

[0081] If the impact factor is equal to 0, it means that there is no obvious linear relationship between the change in temperature and the change in power, and the change in temperature is not obvious enough to affect the change in output power.

[0082] The numerator part of this formula is the covariance of the change rate of the th high-frequency component of the historical temperature at all historical moments and the change rate of the th high-frequency component of the historical power at all historical moments, which reflects the synchronous change trend between the change rate of the high-frequency component of the historical temperature and the change rate of the high-frequency component of the historical power. The larger the covariance, the better the synchronous change trend, and the larger the impact factor. The smaller the covariance, the worse the synchronous change trend, and the smaller the impact factor.

[0083] The denominator part of this formula is the variance of the change rate of the th high-frequency component of the historical temperature at all historical moments, which measures the fluctuation amplitude of the high-frequency component of the historical temperature. The larger the variance, the larger the fluctuation amplitude. Usually, when calculating the covariance and variance, both need to be divided by the total number of samples, that is, the numerator part of this formula needs to be divided by , and the denominator part also needs to be divided by , so the in the numerator and denominator parts cancel each other out.

[0084] Overall, when the temperature change is relatively stable (the denominator is smaller), if the change trend of power is strongly synchronized with the temperature change (the numerator is larger), it indicates that any small change in temperature will have a greater impact on the change in power, and the influence factor value will be larger. Conversely, if the temperature changes violently (the denominator is larger), then even if the synchronization effect between temperature and power is good (the numerator is larger), this synchronization trend will be diluted. Because the violent fluctuation of temperature makes the impact of power fluctuation less prominent, and the value of the influence factor is relatively small.

[0085] In summary, the influence factor calculated by this formula reveals the relationship strength between temperature and power, reflects the actual impact of temperature change on the output power of the photovoltaic power station, and further provides a theoretical basis for power prediction and system optimization.

[0086] In one embodiment, the influence factor of each low-frequency component of historical light intensity and historical temperature on each low-frequency component of historical output power is determined based on the following method: Select any low-frequency component of historical light intensity or historical temperature as the first low-frequency component, and any low-frequency component of historical output power as the second low-frequency component; Take the sum of the values of the first low-frequency component and the second low-frequency component at each historical moment and all previous moments as the cumulative value of the first low-frequency component and the second low-frequency component at that historical moment; Calculate the covariance of the cumulative value of the first low-frequency component at all historical moments and the cumulative value of the second low-frequency component at all historical moments, and the variance of the cumulative value of the first low-frequency component at all historical moments; take the ratio of the covariance and the variance as the influence factor of the first low-frequency component on the second low-frequency component.

[0087] Specifically, the calculation method of the influence factor of each low-frequency component of historical light intensity on each low-frequency component of historical power is as follows:

[0088] In this formula, is the influence factor of the th low-frequency component of historical light intensity on the th low-frequency component of historical power, is the cumulative value of the th low-frequency component of historical power at the th historical moment, and this cumulative value is equal to the sum of the value of the th low-frequency component at the th moment and the accumulated values of the th low-frequency component at all previous moments of the th historical moment, which reflects the The cumulative effect of a low-frequency component over a period of time, and the cumulative value can better capture the progressive relationship between light and power over a longer time range. is the total number of historical moments, is the th low-frequency component of the historical light intensity at the th historical moment, which is equal to the value of the th low-frequency component at the th historical moment plus the sum of the values of the th low-frequency component at all historical moments before the th historical moment. It reflects the cumulative effect of the th low-frequency component of the historical light intensity over a period of time. is the mean value of all the cumulative values generated by the th low-frequency component of the historical power at the th historical moment (since each historical moment generates a cumulative value), reflecting the average level of the low-frequency component of the historical power over a period of time. is the mean value of all the cumulative values generated by the th low-frequency component of the historical light intensity at the th historical moment, reflecting the average level of the low-frequency component of the historical light intensity over a period of time.

[0089] If , it indicates a positive correlation between the low-frequency component of the historical light intensity and the low-frequency component of the historical power, that is, when the light increases, the power also tends to increase, which conforms to the basic physical property that the power rises with the increase of light in the photovoltaic system; if , it means a negative correlation between the two, which may imply situations such as aging or long-term occlusion of the photovoltaic modules. If , it shows that there is no correlation between the low-frequency component of the historical light intensity and the low-frequency component of the historical power.

[0090] In this formula, the numerator is the covariance between the cumulative value of the th low-frequency component of the historical light intensity at all historical moments and the cumulative value of the th low-frequency component of the historical power at all historical moments, and the denominator is the variance of the cumulative value of the th low-frequency component of the historical light intensity at all historical moments.

[0091] In this formula, the numerator measures the long-term synchronous change trend of the cumulative value corresponding to the low-frequency component of the historical light intensity and the cumulative value corresponding to the low-frequency component of the historical power through covariance. If the covariance is larger, it indicates that all the cumulative values corresponding to these two low-frequency components tend to rise or fall simultaneously over a long period, and the long-term synchronous change trend is more consistent, resulting in a larger influence factor; conversely, the same principle applies.

[0092] In this formula, the denominator measures the fluctuation amplitude of the cumulative value corresponding to the low-frequency component of the historical light intensity through variance. Usually, when calculating covariance and variance, it is necessary to divide by the total number of samples. That is, the numerator part of this formula needs to be divided by , and the denominator part also needs to be divided by , so the of the numerator and denominator parts cancel each other out.

[0093] Overall, if the numerator is larger and the denominator is smaller, it indicates that over a long period, the stable increase in light leads to a stable increase in power, indicating that the influence of light on the output power is more obvious. If the numerator is smaller and the denominator is larger, it indicates that over a long period, the change in light fluctuates greatly, and the synchronous change trend between the light intensity and the output power is poor, indicating that the influence of light on the output power is not obvious, suggesting that there are long-term faults in the photovoltaic modules.

[0094] In summary, this formula reasonably captures the long-term relationship between light and power, which helps to evaluate the system health and power prediction of photovoltaic power plants.

[0095] Specifically, the calculation method of the influence factor of each low-frequency component of historical temperature on each low-frequency component of historical power is as follows:

[0096] In this formula, is the influence factor of the -th low-frequency component of historical temperature on the -th low-frequency component of historical power. is the cumulative value of the -th low-frequency component of historical power at the -th historical moment. This cumulative value is equal to the sum of the value of the -th low-frequency component at the -th moment and the sum of the values of the -th low-frequency component at all moments before the -th historical moment. It reflects the cumulative effect of the -th low-frequency component of historical power over a period of time. The cumulative value can better capture the gradual relationship between temperature and power over a longer time range. is the total number of historical moments, is the The cumulative value of the th low-frequency component at the th historical moment is equal to the value of the th low-frequency component at the th historical moment and the cumulative sum of the values of the th low-frequency component at all historical moments before the th historical moment. It reflects the cumulative effect of the th low-frequency component of the historical temperature over a period of time. The mean value of all the cumulative values generated by the th low-frequency component of the historical power at the th historical moment (since each historical moment generates a cumulative value) reflects the average level of the low-frequency component of the historical power over a period of time. The mean value of all the cumulative values generated by the

[0097] If , it indicates that there is a positive correlation between the low-frequency component of the historical temperature and the low-frequency component of the historical power, that is, when the temperature increases, the power also tends to increase, which does not conform to the physical characteristics of the photovoltaic module, meaning that the increase in temperature is positively correlated with the increase in power, which is usually abnormal. This situation may be due to measurement errors or sensor failures. For example, errors in the current sensor, drift or failure of the current sensor when working at high temperatures may cause the measured current value to be too high, resulting in a calculated power value larger than the actual value, or abnormal temperature measurement from the sensor leading to a temperature value larger than the actual value. If , it means that the two are negatively correlated, which conforms to the characteristic that the output power of the photovoltaic module decreases with the rise and fall of temperature. If , it means that there is no correlation between the two.

[0098] The design logic explanation of the numerator and denominator parts of this formula is the same as the description of the corresponding position in the calculation method of the influence factor of the th low-frequency component of the historical light intensity on the th low-frequency component of the historical power, and the only difference is that one is the historical light intensity and the other is the historical temperature.

[0099] S224: Determine the light influence factor and temperature influence factor through weighted average operation.

[0100] First, set the weights of each high-frequency component and low-frequency component: The significance and effect of the weights lie in reasonably evaluating the relative importance of each high-frequency component and low-frequency component in the impact of light and temperature on power by considering the volatility (i.e., variance) of these components at all historical moments. Specifically: The weights are calculated based on the variances of the respective components (the values corresponding to all historical moments), and the variance reflects the volatility or instability of the data. A larger variance means that the component varies more in the time series and may have a more significant impact on power, so it is assigned a larger weight. Conversely, a component with a smaller variance changes less and may have a weaker impact on power, so its weight is smaller.

[0101] In one embodiment, the weight of each high-frequency component of the historical light intensity is the variance of the values of that high-frequency component at all historical moments, divided by the sum of the variances of all high-frequency components and all low-frequency components of the historical light intensity at all historical moments.

[0102] Specifically, the weight of each high-frequency component of the historical light intensity is:

[0103] In this formula, is the weight of the th high-frequency component of the historical light intensity, is the variance of the values of the th high-frequency component of the historical light intensity at all historical moments, is the sum of the variances of all high-frequency components of the historical light intensity at all historical moments, is the variance of the values of the th low-frequency component of the historical light intensity at all historical moments, is the total number of high-frequency components of the historical light intensity, is the sum of the variances of all low-frequency components of the historical light intensity at all historical moments, is the total number of low-frequency components of the historical light intensity.

[0104] In one embodiment, the weight of each low-frequency component of the historical light intensity is the variance of the values of that low-frequency component at all historical moments, divided by the sum of the variances of all high-frequency components and all low-frequency components of the historical light intensity at all historical moments.

[0105] Specifically, the weight of each low-frequency component of the historical light intensity is:

[0106] In this formula, is the weight of the th low-frequency component of the historical light intensity, is the weight of the The variance of the values of the high-frequency components at all historical moments, is the total number of high-frequency components of the historical light intensity, is the sum of the variances of the values of all high-frequency components of the historical light intensity at all historical moments, For the th low-frequency component of the historical light intensity, the variance of the values at all historical moments, is the sum of the variances of the values of all low-frequency components of the historical light intensity at all historical moments, is the total number of low-frequency components of the historical light intensity.

[0107] In one embodiment, the weight of each high-frequency component of the historical temperature is the variance of the values of the high-frequency component at all historical moments, divided by the sum of the variances of the values of all high-frequency components and all low-frequency components of the historical temperature at all historical moments.

[0108] Specifically, the weight of each high-frequency component of the historical temperature is:

[0109] In this formula, is the weight of the th high-frequency component of the historical temperature, is the variance of the values of the th high-frequency component of the historical temperature at all historical moments, is the sum of the variances of the values of all high-frequency components of the historical temperature at all historical moments, is the variance of the values of the th low-frequency component of the historical temperature at all historical moments, is the sum of the variances of the values of all low-frequency components of the historical temperature at all historical moments, is the total number of low-frequency components of the historical temperature, is the total number of high-frequency components of the historical temperature.

[0110] In one embodiment, the weight of each low-frequency component of the historical temperature is the variance of the values of the low-frequency component at all historical moments, divided by the sum of the variances of the values of all low-frequency components of the historical temperature at all historical moments.

[0111] Specifically, the weight of each low-frequency component of the historical temperature:

[0112] In this formula, is the weight of the th low-frequency component of the historical temperature, is the variance of the values of the th low-frequency component of the historical temperature at all historical moments, is the sum of variances of the values of all high-frequency components of historical temperature at all historical moments, is the variance of the values of the -th low-frequency component of historical temperature at all historical moments, is the sum of variances of the values of all low-frequency components of historical temperature at all historical moments, is the total number of low-frequency components of historical temperature, is the total number of high-frequency components of historical temperature.

[0113] Finally, a weighted average operation is performed based on the weights and influence factors of each high-frequency component and each low-frequency component to obtain the light influence factor and the temperature influence factor.

[0114] In one embodiment, the method for obtaining the light factor is as follows: Perform a weighted sum of the weights of each high-frequency component of historical light intensity and the influence factors of each high-frequency component of historical light intensity on each high-frequency component of historical power to obtain the first light accumulation value; perform a weighted sum of the weights of each low-frequency component of historical light intensity and the influence factors of each low-frequency component of historical light intensity on each low-frequency component of historical power to obtain the second light accumulation value; use the sum of the first light accumulation value and the second light accumulation value as the light influence factor.

[0115] The specific calculation formula is:

[0116] In this formula, is the light influence factor, is the weight of the -th high-frequency component of historical light intensity, is the weight of the -th low-frequency component of historical light intensity. is the influence factor of the -th high-frequency component of historical light intensity on the -th high-frequency component of historical power, is the influence factor of the -th low-frequency component of historical light intensity on the -th low-frequency component of historical power, , are respectively the total number of high-frequency components and the total number of low-frequency components of historical light intensity, , are respectively the total number of high-frequency components and the total number of low-frequency components of historical power.

[0117] In this formula, is the average influence factor of all high-frequency components of historical light intensity on all high-frequency components of historical power, representing the overall influence of the instantaneous change of light intensity on the output power. is the average influence factor of all low-frequency components of historical light intensity on all low-frequency components of historical power, representing the overall influence of the long-term trend of light intensity on the output power.

[0118] In one embodiment, the method for obtaining the temperature influence factor is as follows: Perform a weighted sum of the weights of each high-frequency component of historical temperature and the influence factors of each high-frequency component of historical temperature on each high-frequency component of historical power to obtain a first temperature accumulation value; Perform a weighted sum of the weights of each low-frequency component of historical temperature and the influence factors of each low-frequency component of historical temperature on each low-frequency component of historical power to obtain a second temperature accumulation value; Take the sum of the first temperature accumulation value and the second temperature accumulation value as the temperature influence factor.

[0119] The specific calculation formula is:

[0120] In this formula, is the temperature influence factor, is the weight of the th high-frequency component of historical temperature, is the weight of the th low-frequency component of historical temperature, is the th high-frequency component of historical temperature on the th high-frequency component of historical power, is the th low-frequency component of historical temperature on the th low-frequency component of historical power, and are the total number of high-frequency components and the total number of low-frequency components of historical temperature respectively, and are the total number of high-frequency components and the total number of low-frequency components of historical power respectively.

[0121] In this formula, is the average influence factor of all high-frequency components of historical temperature on all high-frequency components of historical power, representing the overall influence of the rapid fluctuation of temperature on the output power.

[0122] In this formula, is the average influence factor of all low-frequency components of historical temperature on all low-frequency components of historical power, representing the overall influence of the long-term change trend of temperature on the output power.

[0123] Through this weighting method, the finally obtained light and temperature influence factors can more accurately reflect the actual influence degrees of high-frequency and low-frequency components at different time scales. Components with higher weights will have a greater impact on the final influence factor, thereby effectively capturing the long-term and short-term effects of light and temperature changes on power.

[0124] S23: Use the light influence factor and the temperature influence factor to correct the preliminary prediction result of the output power at the next moment.

[0125] Since the preliminary prediction result usually has difficulty in accurately capturing the instantaneous influence of high-frequency environmental variables on power output, in this step, the light influence factor and the temperature influence factor are used to correct the preliminary prediction result of the output power at the next moment. By quantifying the dynamic correlation between the high-frequency environmental components and the high-frequency power components, the transient influence ignored in the preliminary prediction is compensated, and the response accuracy of the prediction model to complex environmental changes is improved.

[0126] In one embodiment, a method for correcting the preliminary prediction result of the output power at the next moment is based on the following formula:

[0127] In the formula, is the predicted value of the power after correction at the next moment of the current moment, is the predicted value of the output power at the next moment of the current moment, is the light influence factor, is the temperature influence factor, is the predicted value of the light intensity at the next moment of the current moment, is the predicted value of the temperature at the next moment of the current moment, is the mean value of the historical light intensity at the current moment, is the mean value of the historical temperature at the current moment.

[0128] In this formula, is the light correction term, reflects the characteristic that when the light intensity deviates from the historical mean value, the output power is linearly and positively correlated and adjusted according to the influence factor, which conforms to the linear law of the photovoltaic effect, The larger, the larger, and the power is approximately proportional to the light (when the light intensity is lower than the saturation value).

[0129] In this formula, is the temperature correction term, which reflects the characteristic that when the temperature deviates from the historical mean value, the power is linearly and negatively correlated and adjusted according to the influence factor, which conforms to the linear law of the photovoltaic effect, The larger, The smaller it is, the more it reflects the physical law that the power decreases with the increase in temperature, which is consistent with the negative temperature coefficient characteristic of the component.

[0130] In one embodiment, another method for correcting the preliminary prediction result of the output power at the next moment is as follows: Considering the light saturation effect, when the light intensity exceeds 1000 (watts per square meter), the output power response tends to saturate, and the linear correction will overestimate the gain. The additional light intensity has less and less gain on the power. It is the light intensity saturation value; at the same time, considering the physical material characteristics of the photovoltaic module, the component efficiency attenuation intensifies at a certain temperature critical value, and the linear model cannot reflect the non-linear change of the correction coefficient.

[0131] Therefore, a light saturation effect correction term and a temperature attenuation correction term are introduced and corrected based on the following formula:

[0132] In this formula, is the predicted value of the power corrected at the next moment of the current moment, is the predicted value of the output power at the next moment of the current moment, is the light influence factor, is the temperature influence factor, is the predicted value of the light intensity at the next moment of the current moment, is the predicted value of the temperature at the next moment of the current moment, is the average value of the historical light intensity at the current moment, is the average value of the historical temperature at the current moment, is the light saturation effect correction term, which is used to correct the influence of light intensity on the output power after the light intensity exceeds the saturation point. is the temperature attenuation correction term, which is used to correct the influence when the temperature exceeds the temperature critical value.

[0133] For , the calculation formula is:

[0134] In this formula, is the light intensity saturation value, is the predicted value of the light intensity at the next moment of the current moment, is the natural exponential function. When (when the light intensity is not saturated), as gradually approaches , The closer it is to 1, the closer it is to 1, the closer it is to 1, and at this time, the correction is almost carried out according to the original linear correction logic. And when at this time, it will decay, reducing the impact of excessive light on power. The more saturated the light is, the smaller the impact on power is, and the power tends to be stable and will not increase indefinitely.

[0135] For , the calculation formula is:

[0136] In this formula, is the predicted value of the temperature at the next moment of the current moment, is the temperature critical point. When at this time, is very small, then is approximately 1, which means that the relationship between power and temperature is almost linear, and at this time, the correction is basically based on the original linear correction model. When the temperature exceeds at this time, will increase significantly, and then will decrease rapidly, indicating that the attenuation effect of temperature on power is intensified. When the temperature of the photovoltaic module is relatively high, the amplitude of power decline will increase. After the temperature reaches a certain critical value, the decline rate of power will be further intensified. Therefore, the temperature critical value is set. When the temperature is higher than this critical value, the temperature correction term starts to decay non-linearly.

[0137] The temperature critical value of the photovoltaic module is usually -40 °C to 85 °C. Within this temperature range, the photovoltaic module can work normally. Specifically, the standard working temperature of the photovoltaic module is 25 °C, and at this temperature, the output power of the module reaches 100%. When the working temperature exceeds 25 °C, the output power of the module will decrease as the temperature increases. The change characteristics of the temperature and output power of the photovoltaic module are pre-tested. The module is placed in an environment of 25 °C, and the temperature is gradually increased until 85 °C. For every 1 °C increase, a corresponding output power is obtained. Taking the temperature as the abscissa and the power as the ordinate, the change curve of power with temperature is obtained, and then the slope at each point on the curve is calculated. The temperature value corresponding to the minimum negative slope is used as the temperature critical value , because the minimum value of the negative slope means that the decline rate of power suddenly accelerates, indicating that the temperature intensifies the attenuation of power.

[0138] Through the above steps, the obtained predicted value can adapt to the non-stationary characteristics of power data in the photovoltaic scenario, effectively incorporate external environmental factors, and improve the prediction accuracy.

[0139] Monitoring module S3 monitors the PV power station in real time according to the corrected predicted value. Starting from the current moment, it continuously monitors for one minute. If within this one minute, at a certain moment, the actual power value is lower than 90% or higher than 110% of the corrected predicted value (both are empirical values), mark that moment. If the proportion of the marked moments is greater than or equal to 80% (empirical value), it is considered that the output power of the PV power station has an abnormality, and a warning is given in time to prompt the staff to check. During the monitoring process, taking any moment as the current moment, the same monitoring steps are executed to achieve real-time and accurate monitoring of the PV power station.

Claims

1. A remote monitoring system for a photovoltaic power station based on the Internet of Things, characterized in that, Including: A data acquisition module for monitoring the light intensity, temperature, and output power of a photovoltaic module. A data analysis module for determining the output power at the next moment of the current moment, including: Predicting the light intensity, temperature, and output power at the next moment based on the historical light intensity, historical temperature, and historical output power at each moment through a prediction algorithm; obtaining multiple high-frequency components and low-frequency components of the historical light intensity, historical temperature, and historical power through empirical mode decomposition technology. Obtaining the influence factors of each high-frequency component of the historical light intensity and historical temperature on each high-frequency component of the historical output power, and the influence factors of each low-frequency component of the historical light intensity and historical temperature on each low-frequency component of the historical output power, and determining the light influence factor and temperature influence factor through weighted average operation. Correcting the predicted value of the output power at the next moment according to the predicted values of the light intensity and temperature at the next moment, and the light influence factor and temperature influence factor. A monitoring module for remotely monitoring the operating state of a photovoltaic power station according to the difference between the corrected predicted value and the actual value of the output power at the next moment.

2. The remote monitoring system for a photovoltaic power station based on the Internet of Things according to claim 1, characterized in that, The influence factor of each high-frequency component of the historical light intensity and historical temperature on each high-frequency component of the historical output power is determined based on the following method: Selecting any high-frequency component of the historical light intensity or historical temperature as the first high-frequency component, and any high-frequency component of the historical output power as the second high-frequency component. Calculating all the change rates of the first high-frequency component and all the change rates of the second high-frequency component respectively according to the numerical values of the first high-frequency component and the second high-frequency component at all historical moments. Calculating the covariance between all the change rates of the first high-frequency component and all the change rates of the first high-frequency component, and the variance of all the change rates of the first high-frequency component, and taking the ratio of the covariance and the variance as the influence factor of the first high-frequency component on the second high-frequency component.

3. The remote monitoring system for a photovoltaic power station based on the Internet of Things according to claim 1, characterized in that, The influence factor of each low-frequency component of the historical light intensity and historical temperature on each low-frequency component of the historical output power is determined based on the following method: Selecting any low-frequency component of the historical light intensity or historical temperature as the first low-frequency component, and any low-frequency component of the historical output power as the second low-frequency component. Taking the sum of the numerical values of the first low-frequency component and the second low-frequency component at each historical moment and all the previous moments as the cumulative value of the first low-frequency component and the second low-frequency component at that historical moment. Calculating the covariance between the cumulative values of the first low-frequency component at all historical moments and the cumulative values of the second low-frequency component at all historical moments, and the variance of the cumulative values of the first low-frequency component at all historical moments; taking the ratio of the covariance and the variance as the influence factor of the first low-frequency component on the second low-frequency component.

4. The remote monitoring system for a photovoltaic power station based on the Internet of Things according to claim 1, characterized in that, The method for determining the light influence factor and temperature influence factor through weighted average operation includes: Performing weighted summation according to the weight of each high-frequency component of the historical light intensity / historical temperature and the influence factor of each high-frequency component of the historical light intensity / historical temperature on each high-frequency component of the historical power to obtain the first light cumulative value / first temperature cumulative value. The weight of each low-frequency component of the historical light intensity / historical temperature and the influence factor of each low-frequency component of the historical light intensity / historical temperature on each low-frequency component of the historical power are weighted and summed to obtain a second light accumulation value / second temperature accumulation value; The sum of the first light accumulation value / first temperature accumulation value and the second light accumulation value / second temperature accumulation value is used as the light influence factor / temperature influence factor.

5. The remote monitoring system for a photovoltaic power station based on the Internet of Things according to claim 4, characterized in that, The weight is determined based on the following method: The weight of each high-frequency component / each low-frequency component of the historical light intensity is the variance of the values of this high-frequency component / low-frequency component at all historical moments, divided by the sum of the variances of all high-frequency components and all low-frequency components of the historical light intensity at all historical moments; The weight of each high-frequency component / each low-frequency component of the historical temperature is the variance of the values of this high-frequency component / low-frequency component at all historical moments, divided by the sum of the variances of all high-frequency components and all low-frequency components of the historical temperature at all historical moments.

6. The remote monitoring system for a photovoltaic power station based on the Internet of Things according to claim 1, characterized in that, A method for correcting the predicted value of the output power at the next moment is based on the following formula: ； In the formula, is the predicted value of the power after correction at the next moment of the current moment, is the predicted value of the output power at the next moment of the current moment, is the light influence factor, is the temperature influence factor, is the predicted value of the light intensity at the next moment of the current moment, is the predicted value of the temperature at the next moment of the current moment, is the mean value of the historical light intensity at the current moment, is the mean value of the historical temperature at the current moment.

7. The remote monitoring system for a photovoltaic power station based on the Internet of Things according to claim 1, characterized in that, Another method for correcting the predicted value of the output power at the next moment is based on the following formula: ； In the formula, is the predicted value of the power after correction at the next moment of the current moment, is the predicted value of the output power at the next moment of the current moment, is the light influence factor, is the temperature influence factor, is the predicted value of the light intensity at the next moment of the current moment, is the predicted value of the temperature at the next moment of the current moment, is the average value of the historical light intensity at the current moment, is the average value of the historical temperature at the current moment, is the light saturation effect correction term, is the temperature attenuation correction term.

8. The remote monitoring system for a photovoltaic power station based on the Internet of Things according to claim 7, wherein, The light saturation effect correction term is calculated based on the following formula: ； In the formula, is the correction term for light saturation effect, is the saturation value of light intensity, is the predicted value of the light intensity at the next moment of the current moment, is the natural exponential function.

9. The remote monitoring system for a photovoltaic power station based on the Internet of Things according to claim 7, wherein, The temperature attenuation correction term is determined based on the following formula: ； In the formula, is the temperature decay correction term, is the predicted value of the light intensity at the next moment of the current moment, is the temperature critical value, is the natural exponential function.

10. The remote monitoring system for a photovoltaic power station based on the Internet of Things according to claim 7, wherein, The method for obtaining the temperature critical value is as follows: Pre-test the output power of the photovoltaic module at different temperatures; With the temperature as the abscissa and the output power as the ordinate, construct a coordinate system to obtain the curve of the output power changing with temperature; Calculate the slope at each point on the fitting curve, and take the temperature value corresponding to the smallest negative slope as the temperature critical value.

Images