A power quality monitoring method and system based on an intelligent fusion terminal

Abstract

This invention relates to the field of power quality monitoring, and more particularly to a power quality monitoring method and system based on an intelligent fusion terminal. The method includes: acquiring historical power quality data from the intelligent fusion terminal and preprocessing it; fitting curves to the preprocessed power quality data, extracting maximum and minimum value sequences from the fitted curves, calculating the periodic probability of a preset number of extreme points within the corresponding time period, and determining the periodic characteristics of the power quality signal; constructing a periodic window, calculating a periodic index based on the consistency of power quality data within the periodic window, analyzing the trend of the periodic index changing with the period length, obtaining the optimal feature points, and determining the optimal period length and periodic band to complete power quality monitoring. This invention effectively addresses the nonlinear load problem of the power grid by dynamically adjusting the periodic window length, spline curve fitting, and periodic index analysis, improving the accuracy and processing efficiency of period identification.

Description

Technical Field

[0001] This invention relates to the field of power quality monitoring. In particular, it relates to a power quality monitoring method and system based on an intelligent fusion terminal. Background Technology

[0002] The power system is a vital infrastructure of modern society, and its stable operation is crucial for ensuring social production and daily life. With the continuous expansion and increasing complexity of the power grid, power quality monitoring has become a key link in ensuring the reliable operation of the power system. In recent years, the application of intelligent integrated terminals in power quality monitoring has gradually attracted attention. These terminals can integrate multiple functions to achieve real-time monitoring and data acquisition of power quality, providing strong support for the stable operation of the power system.

[0003] Accurately identifying the periodic characteristics of power quality data is crucial for ensuring the stable operation of the power system during power quality monitoring. Monitoring the periodic waveform of power data enables the timely detection of anomalies in the power grid, such as voltage fluctuations and harmonic interference, allowing for appropriate adjustments and optimizations. With the increasing complexity of the power grid environment, the periodic waveform of power data is subject to interference from various factors, such as harmonics, interharmonics, and voltage fluctuations, making the accurate identification of the periodic waveform even more critical.

[0004] In power quality monitoring, a large amount of data needs to be processed to identify periodic characteristics, which slows down the response speed of the monitoring system. Traditional methods are easily affected by noise and other interference factors, making it difficult to effectively filter out power quality signals with periodic characteristics, leading to misjudgments or omissions. The periodic waveforms of power in the power system also change dynamically with load changes, equipment operating status, and changes in grid topology. Traditional methods are unable to quickly and accurately identify periodic waveforms in power quality data and cannot dynamically adapt to these changes, failing to meet the requirements of real-time monitoring and rapid response in power quality monitoring. Summary of the Invention

[0005] To address the problems in power quality monitoring where traditional methods suffer from slow response due to processing large amounts of data, susceptibility to noise and interference leading to misjudgments or missed judgments, and difficulty in adapting to dynamically changing power cycle waveforms, thus failing to meet the requirements for real-time monitoring and rapid response, this invention provides solutions in the following aspects.

[0006] In the first aspect, a power quality monitoring method based on a smart fusion terminal includes: acquiring historical power quality data from the smart fusion terminal and preprocessing it; fitting curves to the preprocessed power quality data, extracting maximum and minimum value sequences from the fitted curves, calculating the periodic probability of a preset number of extreme points within the corresponding time period, and determining the periodic characteristics of the power quality signal; constructing a periodic window based on the periodic characteristics, calculating a periodic index based on the consistency of power quality data within the periodic window, analyzing the trend of the periodic index changing with the period length, obtaining the feature point that best reflects the periodic characteristics, determining the optimal period length of the feature point, thereby clarifying the periodic band and completing the accurate identification of the power periodic waveform.

[0007] By fitting the power curve to extract the maximum and minimum value sequences, and calculating the periodic probability within relevant time periods, the periodic characteristics of the power quality signal are initially determined. Further, a periodic window is constructed and the periodic index is calculated. The trend of the periodic index changing with the period length is analyzed to determine the optimal period length and clarify the periodic band. Finally, the accurate identification of the power periodic waveform is achieved. By accurately extracting the key features of the power quality signal and analyzing the changes in periodic characteristics, the accuracy and efficiency of identification are significantly improved, while enhancing adaptability to complex power grid environments and interference factors, meeting the needs of real-time monitoring and rapid response.

[0008] Preferably, the method for calculating the periodic probability includes:

[0009] Matlab was used to perform curve fitting on the preprocessed power quality data to obtain the maximum and minimum value sequences of the power curve. The most recent extreme point at the current time was taken as the starting point, and the time periods corresponding to the maximum and minimum values ​​were extracted forward along the time series by a preset number of times.

[0010] Using the first extreme value of any time period as the target extreme value, calculate the time difference between the next extreme value adjacent to the target extreme value, and construct the power fluctuation vector in the order of extreme value, time difference, extreme value, time difference;

[0011] Taking any time period as the target time period, trace back two time periods, namely the first time period and the second time period. Calculate the similarity of the power fluctuation vector between the target time period and the first time period to obtain the first cycle probability, and calculate the similarity of the power fluctuation vector between the target time period and the second time period to obtain the second cycle probability.

[0012] By using Matlab to perform curve fitting on preprocessed power quality data, the maximum and minimum value sequences of the power curve are accurately extracted. Starting from the most recent extreme point, a predetermined number of time periods corresponding to these extreme points are extracted. By constructing a power fluctuation vector containing extreme values ​​and time differences, the similarity of power fluctuation vectors within different time periods is calculated and compared to derive the first and second period probabilities. This quantitatively assesses the periodic characteristics of the power quality signal, improving not only the accuracy and efficiency of period identification but also enhancing adaptability to complex power grid environments and changing interference factors.

[0013] Preferably, in response to the first cycle probability being greater than or equal to the first preset threshold and the second cycle probability being less than the second preset threshold, the power quality signal is considered to have entered a new cycle.

[0014] If the probability of the first cycle is less than the first preset threshold and the probability of the second cycle is greater than or equal to the second preset threshold, then it is considered that there is no new cycle in the power quality signal.

[0015] If the probability of the first cycle is less than the first preset threshold and the probability of the second cycle is less than the second preset threshold, then it is considered that there is no new cycle in the power quality signal.

[0016] If the probability of the first cycle is greater than or equal to the first preset threshold and the probability of the second cycle is greater than or equal to the second preset threshold, then it is considered that there is no new cycle in the power quality signal.

[0017] By comparing the relationship between the first cycle probability and the second cycle probability and a preset threshold, the presence or absence of periodic characteristics in the power quality signal can be effectively identified, significantly improving the accuracy and reliability of periodic characteristic identification.

[0018] Preferably, the calculation method of the cycle index includes:

[0019] Initialize the period window with an initial length of 1, and set the starting point of the period window to the start time of the time series. Iterate through each time point in the time series, calculate the standard deviation of the power quality data at the corresponding time point in each period, and sum the standard deviations of all time points to obtain the total fluctuation of the power quality data in the entire period. Multiply the reciprocal of the sum of the fluctuation of the power quality data and the constant term by the reciprocal of the period length to obtain the period exponent.

[0020] Preferably, the step of obtaining the optimal period length includes:

[0021] Set the initial length of the periodic window to the minimum unit of 1, and construct the initial periodic window from the starting point of the data sequence. Calculate the periodic index of each data point within the initial periodic window, iterate the periodic window, increment the periodic window length by 1 each time, and recalculate the periodic index within each new periodic window until the periodic window covers the entire data sequence or reaches the preset maximum periodic length limit, forming a periodic index sequence. Select the periodic length corresponding to the first maximum value as the optimal periodic length.

[0022] By setting the initial length of the periodic window and iteratively increasing it, the obtained periodic exponential sequence is calculated and analyzed to determine the optimal period length. This allows for the accurate identification of the best periodic characteristics of power quality signals, while improving computational efficiency and adapting to the characteristics of different data sequences. Consequently, it provides reliable periodic characteristic information for in-depth analysis and practical application of power quality signals.

[0023] Preferably, the method for obtaining the cycle index can also be replaced by:

[0024] Initialize the periodic window with an initial length of 1, and set the starting point of the periodic window to the start time of the time series. Iterate through each time point in the time series, calculate the standard deviation of the power quality data at the corresponding time point in each period, and accumulate the standard deviations of all time points. The accumulated result reflects the total fluctuation of the power quality data in the entire period. Add the sum of the fluctuations of the power quality data to the sum of the constant term and take the reciprocal to obtain the periodic index.

[0025] By initializing the periodic window length to 1 and starting from the beginning of the time series, the standard deviation of the power quality data at each time point in the series is calculated. These standard deviations are accumulated to obtain the sum of the power fluctuations within the period. The reciprocal of the sum of this sum and the sum of the small constants is taken to calculate the periodic index, thereby identifying the periodic characteristics in the power quality signal.

[0026] Preferably, the method for obtaining the optimal period length can also be replaced by:

[0027] Initialize the period window with an initial length of 1, and increment the period length by 1. At each length, use Matlab's spline curve fitting function to construct a mapping function with the period length as the x-axis and the period exponent as the y-axis. Calculate the first and second derivatives of the mapping function. On the second derivative curve, obtain the first minimum point. The position corresponding to the minimum point is the maximum change point. The period length corresponding to the maximum change point is determined as the optimal period length.

[0028] Secondly, a power quality monitoring system based on an intelligent fusion terminal includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the aforementioned power quality monitoring method based on an intelligent fusion terminal is implemented.

[0029] The present invention has the following effects:

[0030] 1. This invention effectively addresses nonlinear load and equipment configuration issues in power grids by dynamically adjusting the periodic window length and utilizing Matlab's spline curve fitting technique. Furthermore, through analysis of the periodic exponential sequence, this method can identify the optimal characteristic points of periodic characteristic changes in the power grid, thereby enhancing its adaptability to power grid fluctuations and the algorithm's robustness.

[0031] 2. This invention ensures data quality through preprocessing steps, rapidly identifies periodic characteristics through periodic probability calculation, and accurately determines the optimal period length through periodic index analysis. This not only improves the accuracy of period identification but also enhances processing efficiency by reducing unnecessary calculations, making the method suitable for real-time monitoring and rapid response applications. Attached Figure Description

[0032] Figure 1 This is a flowchart of steps S1-S3 in a power quality monitoring method based on an intelligent fusion terminal according to an embodiment of the present invention.

[0033] Figure 2 This is a structural block diagram of a power quality monitoring system based on an intelligent fusion terminal according to an embodiment of the present invention. Detailed Implementation

[0034] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0035] Reference Figure 1 A power quality monitoring method based on a smart fusion terminal includes steps S1-S3, as detailed below:

[0036] S1: Obtain historical power quality data from the smart converged terminal and perform preprocessing.

[0037] The intelligent fusion terminal collects electrical parameters such as current and voltage through sensors such as current transformers and voltage transformers. These sensors convert analog signals into digital signals, which are then transmitted to the data acquisition module. This invention performs quality monitoring on different types of power quality data.

[0038] Preprocessing includes: cleaning the power quality data to remove outliers and fill in missing values ​​to ensure data integrity; removing noise by applying filtering or wavelet denoising methods; normalizing or standardizing the power quality data to unify the data scale; adjusting the sampling rate as needed; and removing long-term trends to focus on periodic characteristic changes.

[0039] To further explain, when collecting data from multiple devices or multiple locations on the same device simultaneously, it is necessary to ensure the synchronization of data collection in order to facilitate subsequent data integration and analysis.

[0040] In power systems, nonlinear loads such as frequency converters, rectifiers, and switching power supplies inject harmonic currents into the power grid periodically due to the nonlinear relationship between current and voltage. For example, the unidirectional conduction of diodes in a rectifier circuit causes discontinuous current and generates characteristic harmonics such as the 5th and 7th harmonics, based on the theoretical results of Fourier analysis. Furthermore, the high-frequency switching (kHz level) of power electronic devices such as IGBTs (Insulated Gate Bipolar Transistors) and inverters can induce voltage spikes and high-frequency oscillations, forming periodic glitches or high-frequency harmonics. Inappropriate equipment configurations, such as capacitor parameter mismatch or neutral line current imbalance, can exacerbate the harmonic superposition effect, affecting the accuracy of the power cycle waveform.

[0041] Directly calculating the periodic waveform of power quality signals is not only computationally intensive but may also yield inaccurate results. Therefore, preliminary screening is necessary to reduce computational load and confirm the periodic characteristics of real-time power quality signals. The specific steps for quickly identifying the periodic characteristics of power quality signals using an intelligent screening method are as follows:

[0042] S2: Fit curves to the preprocessed power quality data, extract the maximum and minimum value sequences from the fitted curves, calculate the periodic probability of a preset number of extreme points in the corresponding time period, and determine the periodic characteristics of the power quality signal.

[0043] Periodic probabilities include:

[0044] Matlab was used to perform curve fitting on the preprocessed power quality data to obtain the maximum and minimum value sequences of the power curve. The most recent extreme point at the current time was taken as the starting point, and the time periods corresponding to the maximum and minimum values ​​were extracted forward along the time series by a preset number of times.

[0045] Using the first extreme value of any time period as the target extreme value, calculate the time difference between the next extreme value adjacent to the target extreme value, and construct the power fluctuation vector in the order of extreme value, time difference, extreme value, time difference;

[0046] Taking any time period as the target time period, trace back two time periods, namely the first time period and the second time period. Calculate the similarity of the power fluctuation vector between the target time period and the first time period to obtain the first cycle probability, and calculate the similarity of the power fluctuation vector between the target time period and the second time period to obtain the second cycle probability.

[0047] Calculating the time interval between two consecutive extreme values ​​reflects the time span between the two extremes, a key feature for periodic characteristic analysis, because a significant characteristic of periodic signals is the existence of fixed or approximately fixed time intervals. The power fluctuation vector can capture the variation characteristics of the power quality signal within one period.

[0048] To further explain, a predetermined number of time periods corresponding to the maximum and minimum values ​​are extracted from the time series, for example, time points containing two maximum and two minimum values. The purpose of this extraction is to capture the variation characteristics of the power quality signal within a complete cycle, and the data contained in this time period is sufficient to reflect the periodic behavior of the power quality signal.

[0049] By calculating the first and second period probabilities, the periodic characteristics of power quality signals are analyzed in depth, thereby identifying and evaluating the behavioral patterns of power quality signals across different time periods. The first period probability measures the similarity between the current time period and the immediately preceding time period in terms of power fluctuation patterns. Higher similarity indicates consistency in power behavior between these two adjacent time periods, reflecting the stability of the periodic signal. The second period probability measures the similarity between the current time period and an even earlier time period. Lower similarity indicates variability in power behavior over a longer time span, helping to discover the evolution of periodic characteristic patterns. Combining these two probability values ​​allows for more accurate identification of the periodic characteristics of power quality signals and analysis of the consistency and trends of periodic waveforms. This is crucial for applications such as power system monitoring, equipment status analysis, and electrical fault diagnosis.

[0050] Determining the periodic characteristics of power quality signals includes:

[0051] If the probability of the first cycle is greater than or equal to a first preset threshold, and the probability of the second cycle is less than a second preset threshold, then the power quality signal is considered to have entered a new cycle. A high probability of the first cycle indicates that the power fluctuation pattern in the current period is very similar to that of the previous period, which is usually a characteristic of periodic power quality signals. A low probability of the second cycle indicates that the power fluctuation pattern in the current period is significantly different from that of the previous two periods, which means that the power fluctuation pattern in the current period is a new cycle, rather than a repetition of the previous cycle pattern.

[0052] If the probability of the first cycle is less than a first preset threshold and the probability of the second cycle is greater than or equal to a second preset threshold, then it is considered that there is no new cycle in the power quality signal. A low probability of the first cycle indicates that the power fluctuation pattern in the current period is not similar to that of the previous period, while a high probability of the second cycle indicates that the power fluctuation pattern in the current period is similar to that of the previous two periods. This suggests that the power fluctuation pattern in the current period is a repetition of the previous cycle, rather than a new cycle.

[0053] If the probability of the first cycle is less than a first preset threshold and the probability of the second cycle is less than a second preset threshold, then it is considered that there is no new cycle in the power quality signal. Low probabilities of both the first and second cycles indicate that the power fluctuation pattern in the current time period is dissimilar to the previous time period and the two time periods before that. This may be because the power quality signal is subject to interference or significant changes, but these are insufficient to determine a new cycle.

[0054] If the probability of the first cycle is greater than or equal to a first preset threshold, and the probability of the second cycle is greater than or equal to a second preset threshold, then it is considered that there is no new cycle in the power quality signal. A high probability of the first cycle indicates that the power fluctuation pattern in the current time period is very similar to that of the previous time period, while a high probability of the second cycle indicates that the power fluctuation pattern in the current time period is also similar to that of the previous two time periods. This suggests that the power fluctuation pattern in the current time period is a repetition of the previous cycle, rather than a new cycle.

[0055] For example, the first threshold is 0.9 and the second threshold is 0.7, which can be adjusted according to specific circumstances.

[0056] The appearance of a new cycle may indicate an anomaly or fault in the power system. For example, nonlinear characteristics of equipment, interference in the power grid, or sudden changes in load can all cause changes in the periodic characteristics of power quality signals. By detecting the appearance of a new cycle in a timely manner, potential faults or anomalies can be quickly identified, allowing for appropriate measures to be taken for repair or adjustment, thus preventing further damage.

[0057] By identifying new cycles, operating parameters of equipment can be adjusted, load distribution optimized, and system efficiency and reliability improved. Furthermore, analysis of cycle characteristic changes can help predict electricity demand and optimize grid dispatch and management.

[0058] S3: Construct a periodic window based on the periodic characteristics, calculate the periodic index based on the consistency of power quality data within the periodic window, analyze the trend of the periodic index changing with the period length, obtain the feature point that best reflects the periodic characteristics, determine the optimal period length of the feature point, thereby clarifying the periodic band and completing the accurate identification of the power periodic waveform.

[0059] Constructing a periodic window allows us to focus on specific time periods within the power quality data, enabling more accurate analysis of the periodic characteristics of the power quality signal. By iterating through the length of the periodic window and calculating the period exponent and fitting the curve, we can find the optimal period length. This ensures that the analyzed data segment is representative and captures the changing characteristics of the power quality signal within a complete cycle.

[0060] Cyclical indices, including:

[0061] Initialize the period window with an initial length of 1, and set the starting point of the period window to the start time of the time series. Iterate through each time point in the time series, calculate the standard deviation of the power quality data at the corresponding time point in each period, and sum the standard deviations of all time points to obtain the total fluctuation of the power quality data in the entire period. Multiply the reciprocal of the sum of the fluctuation of the power quality data and the constant term by the reciprocal of the period length to obtain the period exponent.

[0062] Specifically, the cyclical index satisfies the following relationship:

[0063] ;

[0064] In the formula, Indicates a cyclical index. Indicates the period length. Indicates the first Within the first cycle Power quality data, This represents the standard deviation function.

[0065] In other words, the cycle length refers to the number of power quality data points contained within a cycle. The smaller the sum of current fluctuation amplitudes at all times within the accumulated cycle, the higher the repetition of power within that cycle. By introducing... This ensures that the cycle length does not increase indefinitely, while maintaining the cycle exponent. The rationality of it.

[0066] Analyzing the trend of the period exponent with the period length is to identify the optimal period length, which best reflects the periodic characteristics of the power quality signal. By observing the trend of the period exponent, the stability and consistency of the power quality signal under different period lengths are determined. The specific steps are as follows:

[0067] Optimal period length includes:

[0068] Set the initial length of the periodic window to the minimum unit of 1, and construct the initial periodic window from the starting point of the data sequence. Calculate the periodic index of each data point within the initial periodic window, iterate the periodic window, increment the periodic window length by 1 each time, and recalculate the periodic index within each new periodic window until the periodic window covers the entire data sequence or reaches the preset maximum periodic length limit, forming a periodic index sequence. Select the periodic length corresponding to the first maximum value as the optimal periodic length.

[0069] The cycle length is increased by incrementing by 1, with each increment yielding one cycle length. This results in a cycle exponential sequence. The first maximum value in the cycle exponential sequence corresponds to the optimal cycle length, which represents the number of power quality data points within the power quality cycle window.

[0070] Set the maximum period window length to When the period length increases to When that time comes, stop calculating the cycle index. The value can be determined based on the actual power quality data sampling frequency and monitoring requirements. For example, for a 50Hz grid frequency, one period is 20ms. If the sampling frequency is 10kHz, there are 500 sampling points in one period. Set it to 500 or adjust it appropriately according to actual needs.

[0071] By finding the period length with the highest period exponent, we can determine that the power quality signal has the strongest periodic characteristics at that length. Analyzing the changing trend of the period exponent allows us to understand the stability of the power quality signal at different period lengths. By selecting an appropriate period length, we can avoid overfitting or underfitting problems caused by excessively long or short period lengths.

[0072] In addition, another embodiment includes:

[0073] Initialize the period window with an initial length of 1, and set the starting point of the period window to the start time of the time series. Iterate through each time point in the time series, calculate the standard deviation of the power quality data at the corresponding time point in each period, and accumulate the standard deviations of all time points. The accumulated result reflects the total fluctuation of the power quality data in the entire period. Add the sum of the fluctuations of the power quality data to the sum of the constant term and take the reciprocal to obtain the period index.

[0074] Initialize the period window with an initial length of 1. Increment the period length sequentially with a step size of 1. At each length, use Matlab's spline curve fitting function to construct a mapping function with the period length as the x-axis and the period exponent as the y-axis. Calculate the first and second derivatives of the mapping function. Obtain the first minimum point on the second derivative curve. The position corresponding to the minimum point is the maximum change point. Determine the period length corresponding to the maximum change point as the optimal period length.

[0075] In Example 2, the period length is not directly limited. Instead, the optimal period length is determined through subsequent fitting and second-order derivative analysis. The latter is more suitable for scenarios with high accuracy requirements and sufficient computing resources.

[0076] The optimal period length was determined, and a period window was constructed based on it. This window length reflects the best periodic characteristics of the power quality signal; therefore, the power quality signal portion within the window is considered a representative periodic waveform. Simply put, the power quality signal portion within the period window is the periodic waveform we identified, enabling dynamic adaptation to different waveform changes and meeting the needs of real-time monitoring and rapid response.

[0077] To further explain, key features can be extracted from the periodic window for further analysis of power quality issues, such as: harmonic content: calculating the amplitude and phase of each harmonic to assess the degree of harmonic pollution; voltage deviation: monitoring the effective value of the voltage to determine whether there are excessively high or low voltage conditions; frequency deviation: measuring the change in grid frequency to ensure it is within the allowable range; the extracted features are then classified to identify different types of power quality problems.

[0078] This invention also provides a power quality monitoring system based on an intelligent fusion terminal. For example... Figure 2 As shown, the system includes a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a power quality monitoring method based on an intelligent fusion terminal according to the first aspect of the present invention. The system also includes other components well-known to those skilled in the art, such as a communication bus and a communication interface; their configurations and functions are known in the art and will not be described further here.

[0079] It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept, and these all fall within the scope of protection of this invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A power quality monitoring method based on an intelligent fusion terminal, characterized in that, include: Acquire historical power quality data from smart converged terminals and perform preprocessing; The preprocessed power quality data is fitted with a curve, and the maximum and minimum value sequences are extracted from the fitted curve. The periodic probability of a preset number of extreme points in the corresponding time period is calculated to determine the periodic characteristics of the power quality signal. The calculation method for the periodic probability includes: Matlab was used to perform curve fitting on the preprocessed power quality data to obtain the maximum and minimum value sequences of the power curve. The most recent extreme point at the current time was taken as the starting point, and the time periods corresponding to the maximum and minimum values ​​were extracted forward along the time series by a preset number of times. Using the first extreme value of any time period as the target extreme value, calculate the time difference between the next extreme value adjacent to the target extreme value, and construct the power fluctuation vector in the order of extreme value, time difference, extreme value, time difference; Taking any time period as the target time period, trace back two time periods, namely the first time period and the second time period. Calculate the similarity of the power fluctuation vector between the target time period and the first time period to obtain the first cycle probability, and calculate the similarity of the power fluctuation vector between the target time period and the second time period to obtain the second cycle probability. A periodic window is constructed based on the periodic characteristics. The periodic index is calculated based on the consistency of power quality data within the periodic window. The trend of the periodic index changing with the period length is analyzed. The feature point that best reflects the periodic characteristics is obtained. The optimal period length of the feature point is determined, thereby clarifying the periodic band and completing the accurate identification of the power periodic waveform. The calculation method for the cycle index includes: Initialize the period window with an initial length of 1, and set the starting point of the period window to the start time of the time series. Iterate through each time point in the time series, calculate the standard deviation of the power quality data at the corresponding time point in each period, and sum the standard deviations of all time points to obtain the total fluctuation of the power quality data in the entire period. Multiply the reciprocal of the sum of the fluctuation of the power quality data and the constant term by the reciprocal of the period length to obtain the period exponent. The steps for obtaining the optimal period length include: Set the initial length of the periodic window to the minimum unit of 1, and construct the initial periodic window from the starting point of the data sequence. Calculate the periodic index of each data point within the initial periodic window, iterate the periodic window, increment the periodic window length by 1 each time, and recalculate the periodic index within each new periodic window until the periodic window covers the entire data sequence or reaches the preset maximum periodic length limit, forming a periodic index sequence. Select the periodic length corresponding to the first maximum value as the optimal periodic length.

2. The power quality monitoring method based on an intelligent fusion terminal according to claim 1, characterized in that, If the probability of the first cycle is greater than or equal to the first preset threshold and the probability of the second cycle is less than the second preset threshold, then the power quality signal is considered to have entered a new cycle. If the probability of the first cycle is less than the first preset threshold and the probability of the second cycle is greater than or equal to the second preset threshold, then it is considered that there is no new cycle in the power quality signal. If the probability of the first cycle is less than the first preset threshold and the probability of the second cycle is less than the second preset threshold, then it is considered that there is no new cycle in the power quality signal. If the probability of the first cycle is greater than or equal to the first preset threshold and the probability of the second cycle is greater than or equal to the second preset threshold, then it is considered that there is no new cycle in the power quality signal.

3. The power quality monitoring method based on an intelligent fusion terminal according to claim 1, characterized in that, The method for obtaining the cycle index can also be replaced by: Initialize the period window with an initial length of 1, and set the starting point of the period window to the start time of the time series. Iterate through each time point in the time series, calculate the standard deviation of the power quality data at the corresponding time point in each period, and accumulate the standard deviations of all time points. The accumulated result reflects the total fluctuation of the power quality data in the entire period. Add the sum of the fluctuations of the power quality data to the sum of the constant term and take the reciprocal to obtain the period index.

4. The power quality monitoring method based on an intelligent fusion terminal according to claim 1, characterized in that, The method for obtaining the optimal period length can also be replaced by: Initialize the period window with an initial length of 1. Increment the period length sequentially with a step size of 1. At each length, use Matlab's spline curve fitting function to construct a mapping function with the period length as the x-axis and the period exponent as the y-axis. Calculate the first and second derivatives of the mapping function. Obtain the first minimum point on the second derivative curve. The position corresponding to the minimum point is the maximum change point. Determine the period length corresponding to the maximum change point as the optimal period length.

5. A power quality monitoring system based on an intelligent fusion terminal, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement the power quality monitoring method based on a smart fusion terminal according to any one of claims 1-4.

Images