Multi-source dynamic power grid harmonic intelligent early warning method and system based on robust optimization

Abstract

The application discloses a multi-source dynamic power grid harmonic intelligent early warning method and system based on robust optimization, relates to the technical field of power system monitoring, and comprises the following specific steps: data acquisition under a multi-source injection scene: through distributed power source monitoring devices, load monitoring devices, and voltage transformers and current transformers, output data of the distributed power source and real-time power data of the load are collected respectively; the application can accurately describe the change rule of the harmonic source by establishing a probability distribution model of the output of the distributed power source and the change of the load; when the harmonic impedance of the power grid is dynamically monitored, the probability distribution model is introduced and different actual operation conditions are simulated, the influence of the uncertainty of the harmonic source on the harmonic impedance is comprehensively considered, the harmonic impedance estimation value after fully considering the uncertainty factors is obtained, dynamic and high-precision monitoring of the change of the harmonic impedance of the power grid with time and operation conditions is realized, and a reliable basis for accurate evaluation of the operation state of the power grid is provided.

Description

Technical Field

[0001] This invention relates to the field of power system monitoring technology, specifically to a robust optimization-based intelligent early warning method and system for multi-source dynamic power grid harmonics. Background Technology

[0002] With the large-scale construction of 750 kV and above AC transmission networks and the widespread application of intelligent dispatching systems, the multi-source injection characteristics of distributed power sources (such as photovoltaic and wind power) and nonlinear loads (such as frequency converters and electric arc furnaces) pose a severe challenge to the security and defense system of large-scale power grids.

[0003] Traditional methods for monitoring and early warning of harmonic impedance have significant limitations. They fail to fully consider the uncertainties of harmonic sources in multi-source injection scenarios. These methods are typically based on fixed operating conditions and parameters, neglecting the randomness and volatility of distributed power output and load changes. Due to the lack of effective handling of harmonic source uncertainties, traditional methods have biases in assessing grid harmonic risks. When harmonic sources fluctuate significantly, the system cannot capture changes in harmonic impedance in a timely and accurate manner, thus failing to issue timely early warning information. This lag and inaccuracy increase the risk of grid failures caused by harmonic problems, leaving the grid without effective monitoring and early warning means when facing complex and ever-changing operating conditions, making it difficult to ensure the safe and stable operation of the grid.

[0004] To address the aforementioned issues, it is necessary to optimize existing methods for dynamic monitoring and intelligent early warning of power grid harmonic impedance. This involves establishing probabilistic models, dynamically monitoring harmonic impedance, designing strategies based on robust optimization theory, and implementing intelligent early warning systems. This will enable high-precision monitoring and reliable early warning of power grid harmonic impedance, ensuring the safe and stable operation of the power grid. Therefore, developing a dynamic monitoring and intelligent early warning technology for harmonic impedance that can comprehensively handle multi-source uncertainties and adapt to the characteristics of power grids above 750 kV is of great significance for improving the security capabilities of large-scale power grids and supporting the efficient operation of intelligent dispatching systems. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a robust optimization-based intelligent early warning method and system for multi-source dynamic power grid harmonics. This method fully considers the uncertainties of harmonic sources in multi-source injection scenarios, establishes a probabilistic model and robust optimization strategy, and comprehensively collects distributed power generation output data, real-time load power data, and voltage and current data of each node in the power grid using various monitoring devices. Through the establishment of a harmonic source uncertainty probabilistic model, it accurately describes the uncertainties of distributed power generation output and load changes. In the harmonic impedance monitoring stage, it dynamically calculates the estimated harmonic impedance considering uncertainties using the probabilistic model, designs monitoring and early warning systems based on robust optimization theory, constructs an optimization model to determine reasonable monitoring parameters and early warning thresholds, and ultimately achieves high-precision monitoring and reliable early warning of power grid harmonic impedance. This ensures the safe and stable operation of the power grid in multi-source injection scenarios and provides strong technical support for the stable operation of the power system.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: On one hand, a robust optimization-based intelligent early warning method for multi-source dynamic power grid harmonics, comprising the following specific steps:

[0007] Data acquisition in multi-source injection scenarios: Through distributed power source monitoring devices, load monitoring devices, voltage transformers, and current transformers, the output data of distributed power sources, real-time power data of loads, and voltage and current data of each node in the power grid are collected respectively.

[0008] Establishment of a probability model for the uncertainty of harmonic sources: Analyze the collected distributed power generation output data and load change data respectively, determine their probability distribution types, estimate the corresponding distribution parameters based on historical data, and establish a probability distribution model that can describe the uncertainty of distributed power generation output and load change.

[0009] Dynamic monitoring of power grid harmonic impedance: Extract the harmonic voltage and current components of each harmonic from the voltage and current data of each node of the power grid, dynamically calculate the harmonic impedance of the power grid at different times according to the definition of harmonic impedance, and integrate the established harmonic source uncertainty probability model into the harmonic impedance calculation process, comprehensively considering the impact of harmonic source uncertainty on harmonic impedance.

[0010] Monitoring and early warning design based on robust optimization theory: An optimization model is constructed with the goals of improving the accuracy of harmonic impedance monitoring and the reliability of early warning. Considering the constraints related to the uncertainty of harmonic sources, the robust optimization algorithm is used to solve the optimization model, and reasonable monitoring parameters and early warning thresholds are determined based on the solution results.

[0011] Intelligent early warning of power grid harmonics: The harmonic impedance calculated by real-time monitoring is compared with the determined early warning threshold. When the harmonic impedance exceeds the early warning threshold, the system issues an early warning message.

[0012] Furthermore, in the step of establishing the probability model for the harmonic source uncertainty, the collected distributed power generation output data and load change data are analyzed to determine their probability distribution types. Based on historical distributed power generation output data, parameter estimation methods are used to determine the specific parameter values ​​corresponding to the distribution type, and a probability distribution model that can characterize the output of distributed power generation is established, thereby reflecting the uncertainty of distributed power generation output. Time series analysis methods are used to analyze the load change data to determine the probability distribution type of the load change data. Based on the characteristics of the load change data itself and the analysis results, a suitable calculation method is used to estimate the relevant parameters of the probability distribution, and a probability distribution model of load change is constructed.

[0013] Furthermore, in the step of establishing the probability model for the harmonic source uncertainty, a probability distribution model capable of characterizing the output of distributed power sources is established, and its model formula is: f(x,y;θ)=f X|Y (x|y;θ X|Y )·f Y (y;θ Y ), where x represents the proportion of distributed power output to rated power, y represents the load power, and θ = (θ X|Y ,θ Y ) represents the set of model parameters, f Y (y;θ Y ) is the marginal probability distribution of load power, and the formula is: Where μ is the mean parameter of the normal distribution, σ is the standard deviation parameter of the normal distribution, and f is calculated from historical load data. X|Y (x|y;θ X|Y ) is the conditional probability distribution of distributed generation output under a given load power y. Considering the spatiotemporal correlation between meteorological factors and load, its formula is: f X|Y in, is the signal variance, which affects the overall variation of the conditional probability distribution; l is the length scale parameter, which controls the degree to which distance affects the correlation in the covariance function. The observation noise variance takes into account the impact of noise present in actual measurements on the model. y is the sample mean of the distributed power output, k(y) is the covariance vector of y and historical load power data, used to measure the correlation between the current load power y and historical data, and K is the covariance matrix, reflecting the correlation structure between historical load power data.

[0014] Furthermore, in the step of establishing the probability model for the uncertainty of the harmonic source, a probability distribution model of load variation is established, and its model parameters are: Where, π k The weight of the k-th mixture component satisfies And π = (π1, π2, ...) follows a Dirichlet distribution DP(α, G0), y represents the load power, and μ k It is the mean parameter of the k-th Gaussian distribution mixture component. It is the variance parameter of the k-th Gaussian distribution mixture component. As a basic distribution, it provides the fundamental distribution form for the generation of mixed components in the Dirichlet process.

[0015] Furthermore, in the step of dynamically monitoring the harmonic impedance of the power grid, the voltage and current components of each harmonic are extracted from the voltage and current data of each node of the power grid. According to the definition of harmonic impedance, the voltage and current components corresponding to each extracted harmonic are calculated. By calculating their ratios, the specific values ​​of each harmonic impedance are obtained. The calculation formula is as follows: Among them, Z h U represents the h-th harmonic impedance. h The h-th harmonic voltage is the voltage value corresponding to the h-th harmonic obtained by processing the collected grid node voltage data. h The h-th harmonic current is the current value corresponding to the h-th harmonic extracted after processing the collected power grid node current data.

[0016] Furthermore, in the step of dynamically monitoring the harmonic impedance of the power grid, the harmonic impedance of the power grid at different times is dynamically calculated according to the definition of harmonic impedance. Specifically, different combinations of distributed power output and load change data that conform to these two probability distributions are randomly generated multiple times. For each data combination, the harmonic impedance under this operating condition is calculated. Through a large number of calculations and analyses, the impact of harmonic source uncertainty on harmonic impedance is comprehensively considered, and the estimated value of harmonic impedance after fully considering the uncertainty factors is obtained, so as to realize the dynamic monitoring of the change of power grid harmonic impedance with time and operating conditions.

[0017] Furthermore, in the step of dynamically monitoring the grid harmonic impedance, for each set of data, the harmonic impedance under this operating condition is calculated. Specifically, N simulations are performed, and each simulation generates a set of distributed power source output P based on the probability model of distributed power source output and load changes. DG,i (i = 1, 2, ..., N) and load power P L The impedance Z of the h-th harmonic is calculated. h,i This yields a harmonic impedance estimate considering uncertainties. Where γ i Z is a correction factor introduced in each simulation based on the current combination of distributed power output and load power, as well as historical operating data of the power grid.h,i This represents the h-th harmonic impedance obtained from the i-th simulation, where N is the simulation order, determined based on the required accuracy and computational resources, according to the basic formula for harmonic impedance. In the i-th simulation, P DG,i and P L,i Under specific operating conditions, the voltage and current data of the power grid nodes are processed to extract the h-th harmonic voltage U. h and current I h Z was calculated h,i , It is the estimated value of the h-th harmonic impedance after considering uncertainties, γ i These are the correction coefficients for each simulation, dynamically generated based on a machine learning model.

[0018] Furthermore, in the step of designing the monitoring and early warning system based on robust optimization theory, an optimization model is constructed with the objectives of improving the monitoring accuracy and early warning reliability of harmonic impedance. Considering the relevant constraints of harmonic source uncertainty, a robust optimization algorithm is used to solve the optimization model, and its formula is as follows: Where J is the objective function value, and w1 and w2 are weighting coefficients, set according to the importance attached to the accuracy of harmonic impedance monitoring and the reliability of early warning, with w1, w2 ≥ 0 and w1 + w2 = 1. The system is tested using these weighting coefficients. During the test, data related to the accuracy of harmonic impedance monitoring and the reliability of early warning are recorded. The recorded data are analyzed. If the monitoring accuracy is found to be below the expected requirement, i.e. If the value of w1 is large, the value of w1 can be increased. The new w1 is calculated using the formula: in These are the weighting coefficients before adjustment, and Δw is the adjustment step size (e.g., Δw = 0.1). H is the total number of harmonic orders considered. It is the estimated h-th harmonic impedance, Z h R is the actual h-th harmonic impedance, and R is the early warning reliability index, which is evaluated based on historical early warning data, and its value ranges from [0, 1]. and These are the minimum and maximum output values ​​of the distributed power source, determined based on its equipment characteristics and operating conditions. DG It is the actual output of the distributed power source. and These are the minimum and maximum values ​​of the load power, determined by the type of load and its operating characteristics. P L This is the actual load power.

[0019] On the other hand, a robust optimization-based multi-source dynamic power grid harmonic intelligent early warning system includes the following components:

[0020] Data acquisition module: used to collect output data of distributed power sources, real-time power data of loads, and voltage and current data of each node in the power grid in multi-source injection scenarios;

[0021] Probability model building module: Based on the data acquired by the data acquisition module, a probability distribution model of the output and load changes of distributed power sources is established to describe the uncertainty of harmonic sources;

[0022] Harmonic impedance monitoring module: Based on the voltage and current data acquired by the data acquisition module, it extracts each harmonic component, calculates the harmonic impedance, and combines the probability model established by the probability model building module to comprehensively consider the impact of harmonic source uncertainty on harmonic impedance.

[0023] Robust Strategy Design Module: Based on robust optimization theory, an optimization model is constructed, the influence of harmonic source uncertainty is considered, reasonable monitoring parameters and early warning thresholds are determined, and robust monitoring and early warning strategies are designed.

[0024] Intelligent early warning module: It compares the real-time harmonic impedance monitored by the harmonic impedance monitoring module with the early warning threshold determined by the robust strategy design module. When the harmonic impedance exceeds the early warning threshold, it issues an early warning message.

[0025] Compared with existing technologies, this robust optimization-based multi-source dynamic power grid harmonic intelligent early warning method and system has the following advantages:

[0026] I. This invention establishes a probability distribution model of distributed power generation output and load changes, which can accurately describe the variation law of harmonic sources. When dynamically monitoring the harmonic impedance of the power grid, these probability models are introduced and different actual operating conditions are simulated. The impact of harmonic source uncertainty on harmonic impedance is comprehensively considered, and the estimated value of harmonic impedance after fully considering the uncertainty factors is obtained. This enables dynamic and high-precision monitoring of the change of power grid harmonic impedance with time and operating conditions, and provides a reliable basis for the accurate assessment of the power grid operating status.

[0027] Second, this invention constructs an optimization model aimed at improving the accuracy of harmonic impedance monitoring and enhancing the reliability of early warning results. It considers various factors related to the uncertainty of harmonic sources as constraints. By solving this model, reasonable monitoring parameters and early warning thresholds are determined, enabling the intelligent early warning module to acquire dynamically monitored harmonic impedance values ​​in real time and accurately compare them with the early warning threshold. When the harmonic impedance exceeds the early warning threshold, it can quickly issue an early warning message according to a pre-set communication method. This effectively solves the problem of not being able to issue timely and accurate early warning messages when harmonic source fluctuations are large, greatly improving the safety and stability of power grid operation.

[0028] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0030] Figure 1 A flowchart of a multi-source dynamic power grid harmonic intelligent early warning method based on robust optimization;

[0031] Figure 2 This is a schematic diagram of a multi-source dynamic power grid harmonic intelligent early warning system based on robust optimization. Detailed Implementation

[0032] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0033] Example 1

[0034] A certain inter-regional power grid connects multiple power generation areas in different regions, including areas where distributed power sources such as large wind farms and photovoltaic power stations are centrally connected, as well as areas with various load types such as industrial clusters, urban commercial and residential mixed power consumption areas. The output of distributed power sources in different regions is significantly uncertain due to the influence of local natural conditions; various loads also exhibit complex and varied power consumption patterns due to differences in regional economic structure and user electricity consumption habits.

[0035] In various distributed power generation areas of cross-regional power grids, such as near each wind turbine in a large wind farm and each photovoltaic array in a photovoltaic power station, high-precision distributed power monitoring devices are installed. These devices, meeting the Nyquist sampling theorem requirement of a 12kHz sampling frequency, collect real-time data on wind speed, wind direction, rotational speed, and real-time power generation of wind turbines, as well as solar irradiance, panel temperature, and real-time power generation of photovoltaic power stations. Simultaneously, the monitoring devices undergo regular maintenance and calibration to ensure the accuracy and reliability of the data.

[0036] In different load areas, including factories in industrial clusters, urban commercial centers, and residential communities, load monitoring devices are rationally deployed according to load distribution and importance, collecting real-time power data of various loads at a sampling frequency of 12kHz. For large industrial users and critical commercial facilities, the density of monitoring points is increased to comprehensively grasp the dynamic changes in load.

[0037] High-precision voltage and current transformers are installed at key nodes in inter-regional power grids, such as at both ends of inter-regional tie lines and busbars of large substations. Voltage and current data from each node are collected at a sampling frequency of 12kHz, and the data is accurately transmitted to the data processing center using well-shielded cables and a reliable data transmission network to prevent signal interference and data loss.

[0038] For distributed generation output data, in-depth analysis is conducted on historical output data collected from large-scale wind farms and photovoltaic power plants. Using data mining and machine learning algorithms, combined with local meteorological data (such as long-term statistical distributions of wind speed and solar irradiance) and characteristic parameters of power generation equipment, a probability distribution model capable of characterizing distributed generation output is established. The model formula is: f(x,y;θ)=f X|Y (x|y;θ X|Y )·f Y (y;θ Y ), where x represents the proportion of distributed power output to rated power, y represents the load power, and θ = (θ X|Y ,θ Y ) represents the set of model parameters, f Y (y;θ Y ) is the marginal probability distribution of load power, and the formula is: Where μ is the mean parameter of the normal distribution, σ is the standard deviation parameter of the normal distribution, and f is calculated from historical load data. X|Y (x|y;θ X|Y ) is the conditional probability distribution of distributed generation output under a given load power y. Considering the spatiotemporal correlation between meteorological factors and load, its formula is: in, is the signal variance, which affects the overall variation of the conditional probability distribution; l is the length scale parameter, which controls the degree to which distance affects the correlation in the covariance function. The observation noise variance takes into account the impact of noise present in actual measurements on the model. y is the sample mean of the distributed power output, k(y) is the covariance vector of y and historical load power data, used to measure the correlation between the current load power y and historical data, and K is the covariance matrix, reflecting the correlation structure between historical load power data.

[0039] For load change data, time series analysis and spectrum analysis methods are used, combined with industry characteristics of loads in different regions, user electricity consumption habits, and other factors, to construct a probability distribution model of load changes. The model parameters are as follows: Where, π k The weight of the k-th mixture component satisfies And π = (π1, π2, ...) follows a Dirichlet distribution DP(α, G0), y represents the load power, and μ k It is the mean parameter of the k-th Gaussian distribution mixture component. It is the variance parameter of the k-th Gaussian distribution mixture component. As a basic distribution, it provides the fundamental distribution form for the generation of mixed components in the Dirichlet process.

[0040] The collected voltage and current data from each node of the power grid are transmitted to the high-performance data processing unit in the data processing center. The Fast Fourier Transform (FFT) algorithm is used to process the data, extracting the voltage and current components of each harmonic. To improve calculation accuracy and reduce spectral leakage errors, the signal is preprocessed according to the definition formula of harmonic impedance. (where Z) h U represents the h-th harmonic impedance. h I represents the h-th harmonic voltage. h Let the current be the h-th harmonic current. Calculate the impedance of each harmonic, introduce a probabilistic model of distributed photovoltaic (PV) output and load variation, and perform 1000 Monte Carlo simulations. In each simulation, the proportion x of the PV power station's output to its rated output is generated based on the probability distribution f(x; α, β) of the PV output, thus obtaining the PV power station's output P. DG P DG =x×P rated ,,P rated Given the rated power of the photovoltaic power station, and based on the probability distribution of load changes, g(y; μ, σ) generates the load power P. L A set of distributed photovoltaic power output P was obtained. D G,i and load power P L The h-th harmonic impedance Z is calculated from i (i = 1, 2, ..., 1000). h,i Combined with the correction factor γ determined based on historical power grid operating data i Through formula Obtain the harmonic impedance estimate considering uncertainties. To enable dynamic monitoring of power grid harmonic impedance.

[0041] An optimization model is constructed with the goals of improving the accuracy of harmonic impedance monitoring and the reliability of early warning. (Where J is the objective function value, which is to be minimized; w1 and w2 are weighting coefficients, initially set to w1 = 0.6 and w2 = 0.4 based on expert experience, with w1, w2 ≥ 0 and w1 + w2 = 1; H is the total number of harmonic orders considered.) It is the estimated h-th harmonic impedance, Z h It is the actual h-th harmonic impedance, and R is the early warning reliability index, which can be evaluated through historical early warning data, etc., with a value range of [0, 1]. The range of distributed photovoltaic output and load changes is taken as a constraint condition, i.e. (in and These are the minimum and maximum output values ​​of a distributed photovoltaic power station, determined based on its equipment characteristics and operating conditions. DG This refers to the actual output of distributed photovoltaic power stations. and These are the minimum and maximum values ​​of the load power, determined by the type of load and its operating characteristics. P L This involves considering the actual load power and the impact of factors such as grid topology and equipment parameters on harmonic impedance. These factors are incorporated into the constraints of the optimization model, and a robust optimization algorithm is used to solve the model. Parallel computing technology is employed to improve computational efficiency during the solution process. The weighting coefficients are adjusted based on experimental test results. Simulation tests are conducted in an actual power grid to collect monitoring accuracy and early warning reliability data under different weighting coefficients. Data analysis and optimization algorithms are then used to determine the optimal weighting coefficients. Specifically, a weighting coefficient running system is used for experimental testing. During the test, data related to harmonic impedance monitoring accuracy and early warning reliability are recorded. The recorded data is analyzed, and if the monitoring accuracy does not meet the expected requirements, [further analysis is needed]. If the value of w1 is large, the value of w1 can be increased. The new w1 is calculated using the formula: in These are the weighting coefficients before adjustment, and Δw is the adjustment step size (e.g., Δw = 0.1). Finally, reasonable monitoring parameters (such as a sampling frequency of 10kHz and a monitoring interval of 5 minutes) and early warning thresholds were determined. In order to improve the accuracy of early warnings, the early warning thresholds were dynamically adjusted and updated in real time according to the operating status of the power grid and load changes.

[0042] The intelligent early warning module acquires the monitored harmonic impedance values ​​in real time. When the harmonic impedance exceeds a predetermined warning threshold, the system notifies the operation and maintenance personnel of the cross-regional power grid via various communication methods (such as SMS, email, and instant messaging software). The notification includes the specific value of the harmonic impedance, the number of times the threshold is exceeded, the possible location of the harmonic source (such as the specific power generation area or load area), and detailed information such as the frequency components of the harmonic. At the same time, an audible and visual alarm is issued on the large screen and professional monitoring software in the power grid monitoring center to attract the attention of the operation and maintenance personnel. The intelligent warning module also provides a detailed harmonic risk analysis report, including an assessment of the impact of harmonics on cross-regional power grid equipment (such as transformers and cables), the possible types of power grid faults (such as equipment overheating and malfunction of protection devices), and targeted countermeasures (such as adjusting the output control strategy of distributed power sources and optimizing load distribution). This helps the operation and maintenance personnel take timely and effective measures to reduce harmonic risks and ensure the safe and stable operation of the cross-regional power grid.

[0043] Example 2

[0044] A wind farm of considerable size is connected to the power grid within an industrial park. This wind farm consists of multiple wind turbine units distributed in specific areas surrounding the park, and is significantly affected by factors such as wind speed, wind direction, and seasonal changes. At the same time, there are a large number of industrial loads within the park, covering multiple industries such as electronics manufacturing, machinery processing, and chemicals. The operation of these industrial loads is highly volatile, and the harmonic characteristics generated by the production equipment in different industries vary. Moreover, their electricity demand changes constantly with production plans and market conditions. The connection of the wind farm makes the already complex grid harmonic situation even more difficult to predict and control, posing a potential threat to the safe operation of power equipment and the stability of power supply within the park.

[0045] Distributed power monitoring devices are installed near each wind turbine in the wind farm. These devices are equipped with high-precision sensors that can accurately measure data such as wind speed, wind direction, turbine speed, and real-time power generation in real time. Data is collected in real time at a sampling frequency of 8kHz. To ensure data continuity and stability, a data redundancy backup mechanism is set up. At the same time, the monitoring devices are regularly inspected on-site and remotely monitored to promptly detect and handle potential equipment failures and data anomalies. In various industrial load areas of the industrial park, load monitoring devices are rationally deployed according to the distribution and importance of the load. For large industrial plants and key production equipment, the density of monitoring points is increased to ensure comprehensive capture of real-time power changes in the load. Real-time power data of industrial loads are also collected at a sampling frequency of 8kHz. Load data from different industries are classified, stored, and managed. High-precision voltage transformers and current transformers are installed at key nodes of the power grid, such as the incoming and outgoing lines of substations and the branch points of important distribution lines. These transformers undergo rigorous calibration and testing to ensure that the measurement accuracy meets the requirements.

[0046] A thorough statistical analysis of wind farm output data is conducted, employing advanced data mining algorithms and signal processing techniques. This is combined with regional meteorological data (such as historical wind speed and direction distribution) and wind farm turbine characteristic parameters (such as rated power and blade size). Probability distribution models are established for different wind turbine models and operating conditions. A comprehensive analysis of industrial load variation data is also performed, integrating time series analysis, spectrum analysis, and other methods. This is combined with the production processes and electricity consumption patterns of various industries. Based on historical load data, statistical inference and parameter estimation methods are used to consider the impact of different time periods (such as weekdays, weekends, and holidays), different production shifts, and changes in market demand on the load. Probability distribution models are established for different types of industrial loads according to industry classification and load characteristics to more accurately describe the uncertainty of industrial load variations.

[0047] The collected voltage and current data are transmitted to the data processing center via a high-speed fiber optic network. This center is equipped with a high-performance server cluster and professional data processing software. The Fast Fourier Transform (FFT) algorithm is used to process the input data, extracting each harmonic component. Based on the definition of harmonic impedance, each harmonic impedance is calculated. A probabilistic model of wind farm output and industrial load changes is introduced, and 800 simulations are performed. During the simulation, the coupling relationship between the wind farm and industrial load, as well as the impact of grid topology changes on harmonic propagation, are considered. Through random sampling and correlation analysis, a reasonable combination of wind farm output and industrial load power data is generated. The corresponding harmonic impedance is calculated for each simulation, and a correction coefficient determined based on historical grid operation and fault records is used. A machine learning-based correction coefficient prediction model is established, comprehensively considering the influence of various factors on the correction coefficient, to obtain an estimate of the harmonic impedance considering uncertainties. This enables dynamic monitoring of the grid's harmonic impedance. Simultaneously, the monitoring results are analyzed and visualized in real time, allowing staff to intuitively understand the grid's harmonic status.

[0048] An optimization model was constructed to improve the accuracy of harmonic impedance monitoring and enhance the reliability of early warning. Weighting coefficients were set (initially w1 = 0.7, w2 = 0.3 based on power grid planning and operational objectives), where w1 focuses on harmonic impedance monitoring accuracy and w2 focuses on early warning reliability. The range of wind farm output and industrial load variations was considered as constraints. Simultaneously, power grid equipment parameters (such as transformer turns ratio and line impedance), operating modes (such as ring network operation and open-loop operation), and the configuration of harmonic mitigation equipment were incorporated into the constraints. A robust optimization algorithm was used to solve the model, employing parallel computing and intelligent optimization. The search strategy improves the efficiency and accuracy of the solution. During the solution process, the weighting coefficients are adjusted in real time according to changes in the power grid operation status, such as new industrial loads, maintenance and repair of wind farm units, etc. By establishing a mapping relationship between the weighting coefficients and the power grid operation parameters, machine learning algorithms are used to adaptively adjust the weighting coefficients. Finally, reasonable monitoring parameters (such as a sampling frequency of 8kHz and a monitoring time interval of 10 minutes) and early warning thresholds are determined. To ensure the scientificity and rationality of the early warning thresholds, the system is verified and optimized by combining simulation and actual testing, taking into account the power grid's safe operation standards and the equipment's tolerance.

[0049] The intelligent early warning module acquires harmonic impedance values ​​in real time and compares them with early warning thresholds. When the harmonic impedance exceeds the threshold, it notifies the power management personnel in the park through system pop-up notifications and high-decibel audible alarms. The system pop-up displays detailed information about the harmonic impedance, the time it exceeded the threshold, relevant harmonic source information, and the potential impact on power grid equipment. The audible alarm uses different frequencies and rhythms to distinguish different levels of harmonic risk. Simultaneously, the intelligent early warning module also sends detailed harmonic risk analysis reports to relevant management personnel via email and instant messaging software. The reports include the frequency components, amplitude, propagation path analysis, potential hazard assessment to equipment, and targeted countermeasures suggestions, helping management personnel to take timely and effective measures to reduce harmonic risks, such as adjusting the operation mode of industrial loads and activating harmonic mitigation equipment.

[0050] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for multi-source dynamic power grid harmonic intelligent early warning based on robust optimization, characterized in that, The method includes the following specific steps: Data acquisition in multi-source injection scenarios: Through distributed power source monitoring devices, load monitoring devices, voltage transformers, and current transformers, the output data of distributed power sources, real-time power data of loads, and voltage and current data of each node in the power grid are collected respectively. Establishment of a probability model for the uncertainty of harmonic sources: Analyze the collected distributed power generation output data and load change data respectively, determine their probability distribution types, estimate the corresponding distribution parameters based on historical data, and establish a probability distribution model that can describe the uncertainty of distributed power generation output and load change. Dynamic monitoring of power grid harmonic impedance: Extract the harmonic voltage and current components of each harmonic from the voltage and current data of each node of the power grid, dynamically calculate the harmonic impedance of the power grid at different times according to the definition of harmonic impedance, and integrate the established harmonic source uncertainty probability model into the harmonic impedance calculation process, comprehensively considering the impact of harmonic source uncertainty on harmonic impedance. Monitoring and early warning design based on robust optimization theory: An optimization model is constructed with the goals of improving the accuracy of harmonic impedance monitoring and the reliability of early warning. Considering the constraints related to the uncertainty of harmonic sources, the robust optimization algorithm is used to solve the optimization model, and reasonable monitoring parameters and early warning thresholds are determined based on the solution results. Intelligent early warning of power grid harmonics: The harmonic impedance calculated by real-time monitoring is compared with the determined early warning threshold. When the harmonic impedance exceeds the early warning threshold, the system issues an early warning message.

2. The method for intelligent early warning of multi-source dynamic power grid harmonics based on robust optimization according to claim 1, characterized in that, In the step of establishing the probability model for the uncertainty of the harmonic source, the collected distributed power generation output data and load change data are analyzed to determine their probability distribution type. Based on historical distributed power generation output data, parameter estimation methods are used to determine the specific parameter values ​​corresponding to the distribution type, and a probability distribution model that can characterize the output of distributed power generation is established to reflect the uncertainty of distributed power generation output. Time series analysis methods are used to analyze the load change data to determine the probability distribution type of the load change data. Based on the characteristics of the load change data itself and the analysis results, a suitable calculation method is used to estimate the relevant parameters of the probability distribution and construct the probability distribution model of the load change.

3. The method for intelligent early warning of multi-source dynamic power grid harmonics based on robust optimization according to claim 1, characterized in that, In the step of establishing the probability model for the harmonic source uncertainty, a probability distribution model capable of characterizing the output of distributed power sources is established, and its model formula is as follows: ,in, This indicates the proportion of distributed power output to rated output. Indicates load power. For the set of model parameters, The marginal probability distribution of load power is given by the formula: ,in, It is the mean parameter of the normal distribution. It is the standard deviation parameter of the normal distribution, calculated from historical load data. At a given load power The conditional probability distribution of distributed power generation output under certain conditions, taking into account the spatiotemporal correlation between meteorological factors and load, is given by the following formula: .

4. The method for intelligent early warning of multi-source dynamic power grid harmonics based on robust optimization according to claim 1, characterized in that, In the step of establishing the probability model for the harmonic source uncertainty, a probability distribution model of load variation is constructed, and its model parameters are as follows: ,in, It is the first The weights of the mixed components satisfy... ,and Follows Dirichlet distribution , Indicates load power. It is the first The mean parameters of a Gaussian-distributed mixture component. It is the first The variance parameter of a Gaussian-distributed mixture component. As a basic distribution, it provides the fundamental distribution form for the generation of mixed components in the Dirichlet process.

5. The method for intelligent early warning of multi-source dynamic power grid harmonics based on robust optimization according to claim 1, characterized in that, In the step of dynamically monitoring the harmonic impedance of the power grid, the voltage and current components of each harmonic are extracted from the voltage and current data of each node of the power grid. According to the definition of harmonic impedance, the voltage and current components corresponding to each extracted harmonic are calculated, and the specific value of each harmonic impedance is obtained by calculating their ratio. The calculation formula is as follows: ,in, express Subharmonic impedance express The subharmonic voltage is obtained by processing and extracting the collected grid node voltage data. The voltage value corresponding to the second harmonic. express The second harmonic current is extracted after processing the collected power grid node current data. The current value corresponding to the subharmonic.

6. The method for intelligent early warning of multi-source dynamic power grid harmonics based on robust optimization according to claim 1, characterized in that, In the step of dynamically monitoring the harmonic impedance of the power grid, the harmonic impedance of the power grid at different times is dynamically calculated according to the definition of harmonic impedance. Specifically, different combinations of distributed power output and load change data that conform to these two probability distributions are randomly generated multiple times. For each data combination, the harmonic impedance under this operating condition is calculated. Through a large number of calculations and analyses, the impact of harmonic source uncertainty on harmonic impedance is comprehensively considered, and the estimated value of harmonic impedance after fully considering the uncertainty factors is obtained, so as to realize the dynamic monitoring of the change of power grid harmonic impedance with time and operating conditions.

7. The method for intelligent early warning of multi-source dynamic power grid harmonics based on robust optimization according to claim 6, characterized in that, In the step of dynamically monitoring the power grid harmonic impedance, for each set of data, the harmonic impedance under this operating condition is calculated. Specifically, let's assume that... Each simulation generates a set of distributed power output values ​​based on a probabilistic model of distributed power generation output and load changes. and load power Calculations yielded Subharmonic impedance This yields a harmonic impedance estimate considering uncertainties. ,in These are correction factors introduced in each simulation based on the current combination of distributed power generation output and load power, as well as historical power grid operation data. Indicates the first The simulation yielded Subharmonic impedance The number of simulations is determined based on the required accuracy and computational resources, according to the basic calculation formula for harmonic impedance. In the The simulation was conducted by and Under specific operating conditions, voltage and current data of power grid nodes are processed and extracted. Subharmonic voltage and current Calculated , It takes uncertainty into account Subharmonic impedance estimate These are the correction coefficients for each simulation, dynamically generated based on a machine learning model.

8. The method for intelligent early warning of multi-source dynamic power grid harmonics based on robust optimization according to claim 5, characterized in that, In the robust optimization theory-based design of the monitoring and early warning steps, an optimization model is constructed with the objectives of improving the monitoring accuracy and early warning reliability of harmonic impedance. Considering the relevant constraints of harmonic source uncertainty, the robust optimization algorithm is used to solve the optimization model, and its formula is as follows: ,in, It is the objective function value. and These are weighting coefficients, set according to the importance placed on the accuracy of harmonic impedance monitoring and the reliability of early warning. , ≥0 and , It is the total number of harmonic orders considered. It is an estimate. Subharmonic impedance It is real. Subharmonic impedance It is an indicator of early warning reliability, evaluated based on historical early warning data, with a value range of [0, 1]. and These are the minimum and maximum output values ​​of the distributed power source, determined based on its equipment characteristics and operating conditions. It is the actual output of the distributed power source. and These are the minimum and maximum values ​​of the load power, determined by the type of load and its operating characteristics. This is the actual load power.

9. A robust optimization-based multi-source dynamic power grid harmonic intelligent early warning system, the system being applicable to the robust optimization-based multi-source dynamic power grid harmonic intelligent early warning method described in any one of claims 1-8, characterized in that, The system includes the following components: Data acquisition module: used to collect output data of distributed power sources, real-time power data of loads, and voltage and current data of each node in the power grid in multi-source injection scenarios; Probability model building module: Based on the data acquired by the data acquisition module, a probability distribution model of the output and load changes of distributed power sources is established to describe the uncertainty of harmonic sources; Harmonic impedance monitoring module: Based on the voltage and current data acquired by the data acquisition module, it extracts each harmonic component, calculates the harmonic impedance, and combines the probability model established by the probability model building module to comprehensively consider the impact of harmonic source uncertainty on harmonic impedance. Robust Strategy Design Module: Based on robust optimization theory, an optimization model is constructed, the influence of harmonic source uncertainty is considered, reasonable monitoring parameters and early warning thresholds are determined, and robust monitoring and early warning strategies are designed. Intelligent early warning module: It compares the real-time harmonic impedance monitored by the harmonic impedance monitoring module with the early warning threshold determined by the robust strategy design module. When the harmonic impedance exceeds the early warning threshold, it issues an early warning message.

Images