A photovoltaic and diesel complementary power supply system in an independent microgrid

By designing microgrid control unit, interference cancellation unit and load optimization distribution unit in independent microgrid, identifying and eliminating clutter interference during photovoltaic and diesel generation, the power supply quality and reliability problems in independent microgrid are solved, and high efficiency is achieved The load distribution and resource allocation have improved the operating efficiency and management level of the microgrid.

CN118713187BActive Publication Date: 2024-12-03JIANGSU LONGQING POWER TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411180512.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-12-03
Estimated Expiration
2044-08-27

AI Technical Summary

Technical Problem

In the prior art, photovoltaic and diesel cannot effectively eliminate clutter interference when power is supplied in an independent microgrid, resulting in electromagnetic interference to other equipment in the microgrid, affecting the power supply quality and reliability.

Method used

A system including microgrid control unit, interference cancellation unit and load optimization distribution unit is designed. Through technologies such as data acquisition, signal simulation, minimum average difference filtering and deep belief network, clutter interference is identified and eliminated, and load optimization is optimized. Allocation to realize effective allocation and coordination of power supply resources.

Benefits of technology

Effectively eliminates clutter interference, improves power supply quality and reliability, optimizes load distribution, reduces energy consumption, improves the operating efficiency and management level of the microgrid, and meets the requirements of sustainable development.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118713187B_ABST
    Figure CN118713187B_ABST
Patent Text Reader

Abstract

The present invention discloses a photovoltaic and diesel complementary power supply system in an independent microgrid, which relates to the field of power grid power supply technology, including a microgrid control unit, an interference elimination unit, a load optimization distribution unit and a monitoring and evaluation unit; the microgrid control unit is used to control the photovoltaic and diesel power supply equipment in the microgrid, and plan and adjust the strategy and state of the microgrid operation; the interference elimination unit is used to estimate and analyze the electric energy data generated during complementary power supply, and eliminate clutter interference using the minimum mean difference method; the load optimization distribution unit is used to establish a prediction model through the complementarity of photovoltaic and diesel power supply to optimize load distribution under deep peak regulation; the monitoring and evaluation unit is used to intelligently monitor the real-time operating state of the independent microgrid and evaluate the quality of power supply. The present invention realizes the efficient configuration and optimization of microgrid resources, effectively improves the overall power supply quality and operating efficiency, and meets the requirements of sustainable development.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of power grid power supply, and in particular to a photovoltaic and diesel complementary power supply system in an independent microgrid. Background Art

[0002] An islanded microgrid refers to a microgrid system that can operate independently from the main grid. Through internal renewable energy generation and energy storage equipment, it can continue to provide power to load users when the main grid is out of power. Generally speaking, compared with microgrids connected to the main grid, independent microgrids have smaller system capacity and closed-loop control and energy management functions. They mainly rely on internal power generation (such as photovoltaics, diesel, etc.) and energy storage equipment to achieve supply and demand balance to ensure stable system operation.

[0003] The photovoltaic and diesel power generation equipment in an independent microgrid have different power generation principles when supplying power. Photovoltaic power generation uses solar energy to directly convert light energy into electrical energy through the photoelectric effect, while diesel power generation generates electricity through an internal combustion engine driving a generator, which requires fuel support. However, the investment cost of photovoltaic power generation is high, but there is no fuel cost. The investment cost of diesel power generation is low, but it needs to bear the fuel cost for a long time.

[0004] Since the load demand of independent microgrids fluctuates greatly, it is difficult for a single energy source to meet it in the long term. In order to ensure the normal power supply, photovoltaic and diesel in the independent microgrid are used for complementary power supply, making full use of the advantages of both, improving the power supply reliability and operating efficiency of the independent microgrid, and meeting the needs of long-term stable power supply.

[0005] However, in the existing technology, when photovoltaic and diesel are used for complementary power supply, it is impossible to effectively eliminate the noise interference generated by the power generation of the two. Since the photovoltaic power generation system uses an inverter to generate pulse current, it is easy to generate high-frequency noise. At the same time, the diesel generator set will also generate low-frequency noise such as mechanical vibration when working. When the two systems are connected to the grid, the noise of photovoltaic and diesel will be superimposed on each other. If effective measures are not taken, these noises are likely to cause electromagnetic interference to other equipment in the microgrid, such as loads, energy storage, etc.

[0006] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0007] In view of the problems in the related art, the present invention proposes a photovoltaic and diesel complementary power supply system in an independent microgrid to overcome the above technical problems existing in the existing related art.

[0008] To this end, the specific technical solution adopted by the present invention is as follows:

[0009] A photovoltaic and diesel complementary power supply system in an independent microgrid, the photovoltaic and diesel complementary power supply system in the independent microgrid comprising a microgrid control unit, an interference elimination unit, a load optimization distribution unit and a monitoring and evaluation unit;

[0010] Wherein, the microgrid control unit is connected to the interference elimination unit, the interference elimination unit is connected to the load optimization distribution unit, and the load optimization distribution unit is connected to the monitoring and evaluation unit;

[0011] Microgrid control unit, used to control the photovoltaic and diesel power supply equipment in the microgrid, and plan and adjust the strategy and status of the microgrid operation;

[0012] An interference elimination unit is used to estimate and analyze the electric energy data generated during complementary power supply, and eliminate clutter interference using the minimum mean difference method;

[0013] The load optimization distribution unit is used to establish a prediction model through the complementarity of photovoltaic and diesel power supply to optimize load distribution under deep peak regulation;

[0014] The monitoring and evaluation unit is used to intelligently monitor the real-time operating status of the independent microgrid and evaluate the quality of power supply.

[0015] Preferably, the interference elimination unit includes a data acquisition and processing module, a clutter determination module, an estimation model construction module, a clutter elimination module and a clutter verification module;

[0016] The data acquisition and processing module is connected to the clutter determination module, the clutter determination module is connected to the estimation model construction module, the estimation model construction module is connected to the clutter elimination module, and the clutter elimination module is connected to the clutter verification module;

[0017] The data acquisition and processing module is used to collect the electric energy data when photovoltaic and diesel are complementary to each other and perform preprocessing operations on it. The electric energy data includes the current and voltage when photovoltaic and diesel are powered, the voltage waveform of the connection point, the input and output current of the inverter when photovoltaic is powered, and the mechanical noise of the internal combustion engine when diesel is powered;

[0018] The clutter determination module is used to establish a simulation model using signal data to determine the clutter signals generated by photovoltaic and diesel power supply, and to classify and extract features of the clutter signals;

[0019] An estimation model building module is used to build a clutter eliminator model using the clutter signal and formulate a filter to estimate the variance of the clutter signal;

[0020] A clutter elimination module is used to eliminate clutter interference by filtering the variance using the minimum mean error;

[0021] The clutter verification module is used to verify whether there is clutter interference in photovoltaic and diesel power supply after clutter elimination.

[0022] Preferably, using signal data to establish a simulation model to determine the clutter signals generated during photovoltaic and diesel power supply, and classifying and extracting features of the clutter signals includes:

[0023] Based on the signal data, the input and output currents of the inverter when photovoltaic power is supplied and the mechanical noise of the internal combustion engine when diesel power is supplied are respectively equivalent to obtain the signal source, which includes a current source and a noise source;

[0024] Combining the vibration, fluctuation, and phase difference mechanism during power supply with the Euler fusion algorithm, a simulation model is constructed, and the process of photovoltaic and diesel complementary power supply is simulated to obtain the phase difference;

[0025] Based on the phase difference, the balance equation is used to calculate the clutter signal in different time periods, and the composition rate of each part of the clutter signal is calculated based on the probability distribution equation. The clutter signal includes voltage fluctuation and harmonics, load change, voltage fluctuation deviation and voltage flicker.

[0026] The clutter signals are classified based on the composition rate, and the features of each group of classified data are extracted by combining the genetic extraction algorithm.

[0027] Preferably, combining the vibration, fluctuation, phase difference mechanism during power supply with the Euler fusion algorithm to build a simulation model, and simulating the process of photovoltaic and diesel complementary power supply includes:

[0028] The simulation model is established by coupling photovoltaic and diesel using discrete dynamics and power transmission principles;

[0029] Input the electric energy data into the simulation model to preset the vibration, fluctuation and phase difference data within different power supply durations;

[0030] The fluctuations and vibrations generated by the simulation are taken as the continuous phase, and the phase difference is taken as the discrete phase. The Euler fusion algorithm is used to track the discrete phase to determine the phase difference generated when the power supply is calculated based on the duration.

[0031] Preferably, the calculation formula of the phase difference is:

[0032] ;

[0033] In the formula, W Represents the phase difference value, represents a constant, e Indicates the current generated by diesel powered equipment, t represents the simulation time point, Indicates the current frequency value of diesel powered equipment, aIndicates the current value generated by the photovoltaic power supply equipment. l Indicates the power supply duration.

[0034] Preferably, constructing a clutter eliminator model using the clutter signal and formulating a filter to estimate the variance of the clutter signal includes:

[0035] Construct a clutter model based on the extracted clutter signal features, design different clutter eliminators based on the clutter model, and establish the elimination capability index and threshold rules when the clutter eliminator is used;

[0036] A clutter eliminator model is established based on the elimination capability index, and the clutter signal is input into the clutter eliminator model to generate adversarial data, and the adversarial data is normalized and the covariance matrix is ​​solved;

[0037] Wavelet coefficients are used to solve the index weights of adversarial data, and the weighted sum method is used to aggregate the indexes, which are used as explicit variables of the clutter eliminator model.

[0038] A clutter eliminator program is written according to the clutter eliminator model, and a maximum estimation method is used to perform calculations to obtain an estimated value of clutter in the clutter eliminator model;

[0039] Determine whether the estimated value meets the threshold rule. If it is within the threshold range, it is retained; if it is not within the threshold range and does not meet the rule, it is discarded;

[0040] Based on the estimated values, the quantitative relationship model between the elimination ability index and the manifest variable is obtained, and the elimination ability is evaluated;

[0041] According to the evaluation results, a wavelet filter for clutter interference is constructed, and the wavelet filter is applied to the clutter signal for filtering to obtain the electric energy signal, and the variance of the electric energy signal is estimated.

[0042] Preferably, using the minimum mean error to filter the variance to eliminate clutter interference includes:

[0043] Determine the window size of the waveform mean and variance of the electric energy signal, and use the minimum mean error method to filter the variance;

[0044] The distribution characteristics of the preset electric energy signal are calculated to obtain the variance compactness information;

[0045] Based on the variance compactness information, candidate initial clusters that meet the clutter interference elimination conditions are screened out, and the candidate initial clusters are processed and screened to obtain variances that meet the elimination conditions;

[0046] The wavelet filter that can eliminate clutter interference is obtained by inverse analysis based on the variance results and is used for photovoltaic and diesel complementary power supply to ensure no clutter interference.

[0047] Preferably, the load optimization distribution unit includes a data collection and processing module, a load prediction module, a deep peak-shaving module and a distribution verification module;

[0048] Among them, the data collection and processing module is connected to the load prediction module, the load prediction module is connected to the deep peak-shaving module, and the deep peak-shaving module is connected to the allocation verification module;

[0049] Data collection and processing module, used to collect historical operating data of photovoltaic and diesel equipment and pre-process the operating data, including power output, load demand, fuel consumption and weather conditions on power supply days;

[0050] The load forecasting module is used to predict the load of complementary power supply based on the operating principle of photovoltaic and diesel complementary power generation and select operating data using an improved deep belief network;

[0051] The deep peak-shaving module is used to build a deep peak-shaving load distribution model based on the load forecast results and solve it to obtain the distribution results;

[0052] The distribution verification module is used to verify the accuracy of load distribution during complementary power supply and optimize the distribution model based on the verification results.

[0053] Preferably, according to the operating principle of photovoltaic and diesel complementary power generation, the load forecasting of complementary power supply using the improved deep belief network by selecting operating data includes:

[0054] According to the principle of photovoltaic power generation, the weather conditions on the power supply day are selected as the prediction data affecting the photovoltaic power supply, and the fuel consumption is selected as the prediction data affecting the diesel power supply;

[0055] The two sets of prediction data are used as the input values ​​of the prediction model, and the improved deep belief network algorithm is used to generate the visible layer and the hidden layer;

[0056] Using multi-layer restricted Boltzmann machine stacking, the load demand is used as the model output value to generate the power supply load forecasting mechanism input to the visual layer;

[0057] The deep features of the input value are extracted and passed to the hidden layer to determine the hidden layer activation function, and the input value is extracted from the activation function to complete the update iteration to obtain the load forecasting model to achieve photovoltaic and diesel complementary power supply in line with the prediction.

[0058] Preferably, building a deep peak load distribution model according to the load forecast results, and solving to obtain the distribution results includes:

[0059] The dynamic allocation parameters of power supply are calibrated according to the prediction results, and the dynamic allocation and power generation balance on the power supply day are preset as the target;

[0060] Based on the naive Bayes method, a combination method is used to express the objective function of the dynamic allocation of power supply on a daily basis;

[0061] On the basis of the objective function, a boundary function is set in combination with the power generation balance target, and the deep peak load distribution model is obtained by summing the objective function and the boundary function.

[0062] The required power generation is input into the allocation model to solve the output allocation results of photovoltaic and diesel complementary power supply during each peak load time.

[0063] The beneficial effects of the present invention are:

[0064] 1. The present invention first estimates and analyzes the complementary power supply data of photovoltaic and diesel power supply in the microgrid, uses the minimum mean difference method to eliminate clutter interference, and establishes a photovoltaic-diesel-load prediction model for deep peak-shaving optimization, thereby realizing the effective configuration and coordination of microgrid power supply resources and improving the overall power supply efficiency. It identifies power supply quality problems through data analysis, and adjusts and controls the equipment in a targeted manner to improve the power supply quality. It can not only eliminate clutter interference and effectively improve the power supply quality and reliability, but also establish a prediction model to predict the supply and demand relationship in advance, perform deep peak-shaving and optimize load distribution, and reduce costs. Through data-driven predictive control means, efficient configuration and optimization of microgrid resources are achieved, effectively improving the overall power supply quality and operating efficiency, and meeting the requirements of sustainable development.

[0065] 2. The present invention is based on establishing a simulation model using signal data, determining the clutter signals generated during photovoltaic and diesel power generation, and performing classification and identification as well as feature extraction, and then constructing a clutter eliminator model, using a filter to estimate the variance of the clutter signal, and using the minimum mean difference method to filter the variance to achieve clutter interference elimination, and finally verifying whether the clutter interference is successfully eliminated. The present invention can identify various types of clutter generated at different stages of photovoltaic and diesel power generation, which is conducive to the targeted design of filters to eliminate different types of clutter, and identify and filter clutter at the digital signal processing level instead of by adding additional equipment, which is more flexible and efficient, and uses variance estimation and minimum mean difference filtering methods, which can effectively reduce the power of clutter, remove clutter interference to the maximum extent, improve the quality and reliability of photovoltaic and diesel power generation, and create conditions for stable power supply of microgrids.

[0066] 3. The present invention improves the deep belief network to predict the complementary power supply load of photovoltaic and diesel, builds a deep peak-shaving load distribution model and solves the distribution results to better learn and express the photovoltaic-diesel complementary operation rules and improve the accuracy of load prediction. Based on the load prediction results, the deep peak-shaving model can achieve efficient matching configuration of supply and demand resources, reduce energy consumption, provide quantitative reference for microgrid operation, guide actual dispatching and control, realize the automatic and efficient configuration of microgrid resources, improve the operation management level, and have positive significance for the operation and management of microgrids. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0068] Figure 1 The present invention is a block diagram of a photovoltaic and diesel complementary power supply system in an independent microgrid according to an embodiment of the present invention.

[0069] In the figure:

[0070] 1. Microgrid control unit; 2. Interference elimination unit; 201. Data acquisition and processing module; 202. Clutter determination module; 203. Estimation model building module; 204. Clutter elimination module; 205. Clutter verification module; 3. Load optimization allocation unit; 301. Data collection and processing module; 302. Load prediction module; 303. Deep peak regulation module; 304. Allocation verification module; 4. Monitoring and evaluation unit. DETAILED DESCRIPTION

[0071] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments. They can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in this field should be able to understand other possible implementation methods and advantages of the present invention.

[0072] According to an embodiment of the present invention, a photovoltaic and diesel complementary power supply system in an independent microgrid is provided.

[0073] The present invention will now be further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to an embodiment of the present invention, a photovoltaic and diesel complementary power supply system in an independent microgrid includes a microgrid control unit 1, an interference elimination unit 2, a load optimization distribution unit 3 and a monitoring and evaluation unit 4.

[0074] Among them, the microgrid control unit 1 is connected to the interference elimination unit 2, the interference elimination unit 2 is connected to the load optimization distribution unit 3, and the load optimization distribution unit 3 is connected to the monitoring and evaluation unit 4.

[0075] The microgrid control unit 1 is used to control the photovoltaic and diesel power supply equipment in the microgrid, and plan and adjust the strategy and status of the microgrid operation.

[0076] In this embodiment, the control of photovoltaic and diesel power supply equipment in the microgrid, and the planning and adjustment of the strategy and status of the microgrid operation mainly include the following aspects:

[0077] According to the photovoltaic power generation forecast, the day-ahead operation plan of the photovoltaic generator set is formulated to determine the expected power generation. According to the load forecast and power generation plan, the operating status of the diesel generator set is planned to determine the start and stop times and power generation.

[0078] Formulate a charging and discharging plan for the energy storage system, coordinate the connection and operation of photovoltaic, diesel and energy storage, and send power generation control signals to the photovoltaic units according to the operation plan, such as maximum power point tracking, and send start and stop, power output and other control instructions to the diesel generator sets to achieve automatic regulation.

[0079] Control the charging and discharging of the energy storage system to ensure power balance, monitor various operating data and status, dynamically adjust the operation plan, achieve closed-loop control, optimize the operation strategy, and improve economy and stability, such as setting multiple operating modes.

[0080] Analyze operating data, evaluate control effects, update control strategies and equipment parameters, and work with the energy management system to automatically adjust the microgrid operating status according to changes in load demand.

[0081] The microgrid control unit 1 can realize coordinated control of photovoltaic and diesel power generation equipment and optimization of microgrid operation strategy, thereby improving system efficiency, stability and economy.

[0082] The interference elimination unit 2 is used to estimate and analyze the electric energy data generated during complementary power supply, and eliminate clutter interference using the minimum mean difference method.

[0083] In this embodiment, the interference elimination unit 2 includes a data acquisition processing module 201 , a clutter determination module 202 , an estimation model construction module 203 , a clutter elimination module 204 and a clutter verification module 205 .

[0084] The data acquisition and processing module 201 is connected to the clutter determination module 202 , the clutter determination module 202 is connected to the estimation model construction module 203 , the estimation model construction module 203 is connected to the clutter elimination module 204 , and the clutter elimination module 204 is connected to the clutter verification module 205 .

[0085] The data acquisition and processing module 201 is used to collect the electric energy data when photovoltaic and diesel complementary power supply is used, and perform preprocessing operations on it. The electric energy data includes the current and voltage when photovoltaic and diesel are used for power supply, the voltage waveform of the connection point, the input and output current of the inverter when photovoltaic power supply is used, and the mechanical noise of the internal combustion engine when diesel power supply is used.

[0086] Specifically, collecting the electric energy data when photovoltaic and diesel are complementary to each other and preprocessing the data includes the following steps:

[0087] Collect the current and voltage when photovoltaic and diesel are powered, the voltage waveform of the connection point, the input and output current of the inverter when photovoltaic is powered, and the mechanical noise of the internal combustion engine when diesel is powered, and perform format verification on the collected data to ensure that the data fields are correct.

[0088] Missing or abnormal data can be marked or repaired, such as large or small values, inconsistent formats, etc., and the data can be aligned at regular intervals to facilitate comparative analysis of data from different devices. The data can be normalized and mapped to a range of dimensions with similar values.

[0089] Resample the data to obtain a uniform time interval sampling sequence, sort the data, and arrange them in chronological order to facilitate subsequent analysis.

[0090] It should be noted that the purpose of preprocessing is to improve data quality and provide reliable data support for subsequent feature extraction, model training, predictive analysis, etc.

[0091] The clutter determination module 202 is used to establish a simulation model using signal data to determine the clutter signals generated during photovoltaic and diesel power supply, and to classify and extract features of the clutter signals.

[0092] Specifically, the signal data is used to establish a simulation model to determine the clutter signals generated by photovoltaic and diesel power supply, and the clutter signals are classified and feature extracted, including:

[0093] Based on the signal data, the input and output currents of the inverter when powered by photovoltaic power and the mechanical noise of the internal combustion engine when powered by diesel power are respectively equivalent to obtain the signal source, which includes a current source and a noise source.

[0094] A simulation model is constructed by combining the vibration, fluctuation, and phase difference mechanism during power supply with the Euler fusion algorithm, and the process of photovoltaic and diesel complementary power supply is simulated to obtain the phase difference.

[0095] Among them, the vibration, fluctuation, phase difference mechanism during power supply and Euler fusion algorithm are combined to build a simulation model, and the process of photovoltaic and diesel complementary power supply is simulated, including:

[0096] The simulation model is established by coupling photovoltaic and diesel using discrete dynamics and power transmission principles;

[0097] Input the electric energy data into the simulation model to preset the vibration, fluctuation and phase difference data within different power supply durations;

[0098] The fluctuations and vibrations generated by the simulation are taken as the continuous phase, and the phase difference is taken as the discrete phase. The Euler fusion algorithm is used to track the discrete phase to determine the phase difference generated when the power supply is calculated based on the duration.

[0099] It should be noted that the calculation formula of the phase difference is:

[0100] ;

[0101] In the formula, W Represents the phase difference value, represents a constant, e Indicates the current generated by diesel powered equipment, t represents the simulation time point, Indicates the current frequency value of diesel powered equipment, a Indicates the current value generated by the photovoltaic power supply equipment. l Indicates the power supply duration.

[0102] It should be explained that the constant in the calculation formula of the phase difference is usually 1. Since the phase itself is an angle value, representing the current position of a sine wave, the phase difference is the difference between the two phase angles, so the difference can be directly calculated. Difference between the two phases is equivalent to rotating the two phases to the same coordinate axis, and then calculating the angle difference. Because no scaling transformation is required, the constant can be 1. This constant 1 will not have a substantial impact on the result. It is mainly for the purpose of unifying the expression format.

[0103] When the Euler fusion algorithm is used to calculate the phase difference, the constant in the formula is usually simply taken as 1, and no special setting is required. This is mainly to express the meaning of performing a difference operation on two phases.

[0104] Based on the phase difference, the balance equation is used to calculate the clutter signal in different time periods, and the composition rate of each part of the clutter signal is calculated based on the probability distribution equation. The clutter signal includes voltage fluctuation and harmonics, load change, voltage fluctuation deviation and voltage flicker.

[0105] Specifically, the clutter signal in different time periods is calculated using a balance equation based on the phase difference, and the composition rate of each part of the clutter signal is calculated based on the probability distribution equation, including the following steps:

[0106] The voltage and current waveform data of different time periods are collected, the phase of the fundamental signal is extracted through FFT and other methods, the phase difference between the fundamental signals at different times is calculated, and a balanced equation model is established. The phase difference is used as the input variable to calculate the clutter signals of different time periods.

[0107] Collect statistical distribution data of clutter signals, such as harmonic content distribution, amplitude distribution, etc., establish a probability distribution equation, use statistical distribution as input, determine the probability density function of various types of clutter, use the total energy of the clutter signal as a normalized constant, and calculate the energy composition ratio of each clutter part.

[0108] The probability distribution equation is used to calculate the clutter in different time periods to obtain the composition rate of each part of the clutter. The changing law and influencing factors of the clutter composition rate in different time periods are analyzed, and the calculation accuracy and effect of the balance equation and probability distribution equation are evaluated.

[0109] The clutter signals are classified based on the composition rate, and the features of each group of classified data are extracted by combining the genetic extraction algorithm.

[0110] Specifically, the clutter signal is classified based on the composition rate, and the feature extraction of each group of classified data is performed in combination with the genetic extraction algorithm, including:

[0111] According to the composition ratio of each component in the clutter signal, the classification threshold is set to divide the clutter signal into different categories, such as harmonic category, phase modulation category, etc. The data of each category is sampled to obtain signal samples, and the individual encoding method of the genetic algorithm is designed. Each individual represents a candidate solution for feature extraction.

[0112] The fitness function of the genetic algorithm can be set. The accuracy of signal classification can be used as the fitness. The genetic algorithm can be executed to generate new feature extraction schemes through selection, crossover, mutation, etc. The individual with the highest fitness is selected from the final population, and its corresponding feature extraction scheme is obtained. The scheme is used to extract features from each category of clutter signal samples.

[0113] The clutter signal classification model is trained using the extracted features, the performance of the classification model is evaluated on the test set, and the optimization is repeated until the classification achieves a satisfactory accuracy.

[0114] In this way, an optimized feature extraction scheme for different clutter signals can be obtained, thereby improving the subsequent classification and recognition performance.

[0115] The estimation model building module 203 is used to build a clutter eliminator model using the clutter signal and formulate a filter to estimate the variance of the clutter signal.

[0116] Specifically, the clutter eliminator model is constructed using the clutter signal, and a filter is formulated to estimate the variance of the clutter signal, including:

[0117] Construct a clutter model based on the extracted clutter signal features, design different clutter eliminators based on the clutter model, and establish the elimination capability index and threshold rules when the clutter eliminator is used;

[0118] A clutter eliminator model is established based on the elimination capability index, and the clutter signal is input into the clutter eliminator model to generate adversarial data, and the adversarial data is normalized and the covariance matrix is ​​solved;

[0119] Wavelet coefficients are used to solve the index weights of adversarial data, and the weighted sum method is used to aggregate the indexes, which are used as explicit variables of the clutter eliminator model.

[0120] A clutter eliminator program is written according to the clutter eliminator model, and a maximum estimation method is used to perform calculations to obtain an estimated value of clutter in the clutter eliminator model;

[0121] Determine whether the estimated value meets the threshold rule. If it is within the threshold range, it is retained; if it is not within the threshold range and does not meet the rule, it is discarded;

[0122] Based on the estimated values, the quantitative relationship model between the elimination ability index and the manifest variable is obtained, and the elimination ability is evaluated;

[0123] According to the evaluation results, a wavelet filter for clutter interference is constructed, and the wavelet filter is applied to the clutter signal for filtering to obtain the electric energy signal, and the variance of the electric energy signal is estimated.

[0124] The clutter elimination module 204 is used to eliminate clutter interference by filtering the variance using the minimum mean error.

[0125] Specifically, using the minimum mean error to filter the variance to eliminate clutter interference includes:

[0126] Determine the window size of the waveform mean and variance of the electric energy signal, and use the minimum mean error method to filter the variance;

[0127] The distribution characteristics of the preset electric energy signal are calculated to obtain the variance compactness information;

[0128] Based on the variance compactness information, candidate initial clusters that meet the clutter interference elimination conditions are screened out, and the candidate initial clusters are processed and screened to obtain variances that meet the elimination conditions;

[0129] The wavelet filter that can eliminate clutter interference is obtained by inverse analysis based on the variance results and is used for photovoltaic and diesel complementary power supply to ensure no clutter interference.

[0130] The clutter verification module 205 is used to verify whether there is clutter interference in the photovoltaic and diesel power supply after the clutter is eliminated.

[0131] Specifically, verifying whether there is clutter interference in photovoltaic and diesel power supply after clutter elimination can mainly include:

[0132] Power quality analyzers are installed at the access points of photovoltaic and diesel power supply systems to collect power quality data such as voltage and current. The collected data are analyzed to check whether indicators such as harmonic content, voltage fluctuation, and flicker meet the power quality standards. Voltage waveform data at the power supply connection point is collected, and the harmonic components of the waveform are analyzed using methods such as FFT.

[0133] Check whether there are sudden waves or inrush currents in the current waveform, record the working conditions of important load equipment when the power supply mode is switched, check whether transient faults occur, count whether the additional losses of motors and transformers caused by harmonics are significant, monitor the stability of the power supply quality over a long period of time, and check whether there are intermittent sudden interferences.

[0134] Compare the power quality data before and after the noise control, analyze the control effect, and take further noise control measures if necessary according to the power quality analysis results, until all indicators meet the standard requirements and it is confirmed that the power supply quality is good, then the noise interference can be considered to be effectively controlled.

[0135] In summary, the present invention is based on establishing a simulation model based on signal data, determining the clutter signals generated during photovoltaic and diesel power generation, and performing classification and identification as well as feature extraction, and then constructing a clutter eliminator model to eliminate clutter interference, and finally verifying whether the clutter interference is successfully eliminated. It can effectively reduce the power of clutter, remove clutter interference to the maximum extent, improve the quality and reliability of photovoltaic and diesel power generation, and create conditions for stable power supply of microgrids.

[0136] The load optimization distribution unit 3 is used to establish a prediction model through the complementarity of photovoltaic and diesel power supply to optimize the load distribution under deep peak regulation.

[0137] In this embodiment, the load optimization distribution unit 3 includes a data collection and processing module 301 , a load prediction module 302 , a deep peak regulation module 303 and a distribution verification module 304 .

[0138] The data collection and processing module 301 is connected to the load prediction module 302 , the load prediction module 302 is connected to the deep peak-shaving module 303 , and the deep peak-shaving module 303 is connected to the allocation verification module 304 .

[0139] The data collection and processing module 301 is used to collect historical operating data of photovoltaic and diesel equipment and pre-process the operating data. The operating data includes power output, load demand, fuel consumption and weather conditions on power supply days.

[0140] Specifically, the preprocessing of historical operation data of photovoltaic and diesel equipment includes:

[0141] Data collection: Collect data such as power generation, sunshine, ambient temperature, etc. on the photovoltaic side; collect data such as power generation, load rate, fuel consumption, etc. on the diesel generator side; collect power consumption and load curve data on the load side; record text information such as operation and maintenance logs in different time periods.

[0142] Data conversion: Convert categorical data into numerical or flag data, and standardize field representation between different tables.

[0143] Data fusion: Merge heterogeneous data from different sources into a unified table, align them with timestamps to ensure matching of data at the same time, arrange data in chronological order, and group and sort by device or type.

[0144] In summary, after preprocessing, a standardized data set is formed, which provides well-structured data input for algorithms such as model training and data mining.

[0145] The load prediction module 302 is used to perform load prediction of complementary power supply based on the operating principle of photovoltaic and diesel complementary power generation and select operating data using an improved deep belief network.

[0146] Specifically, according to the operating principle of photovoltaic and diesel complementary power generation, the load forecasting of complementary power supply using improved deep belief network is carried out by selecting operating data, including:

[0147] According to the principle of photovoltaic power generation, the weather conditions on the power supply day are selected as the prediction data affecting the photovoltaic power supply, and the fuel consumption is selected as the prediction data affecting the diesel power supply;

[0148] The two sets of prediction data are used as the input values ​​of the prediction model, and the improved deep belief network algorithm is used to generate the visible layer and the hidden layer;

[0149] Using multi-layer restricted Boltzmann machine stacking, the load demand is used as the model output value to generate the power supply load forecasting mechanism input to the visual layer;

[0150] The deep features of the input value are extracted and passed to the hidden layer to determine the hidden layer activation function, and the input value is extracted from the activation function to complete the update iteration to obtain the load forecasting model to achieve photovoltaic and diesel complementary power supply in line with the prediction.

[0151] The deep peak-shaving module 303 is used to build a deep peak-shaving load distribution model according to the load forecast results, and solve it to obtain the distribution results.

[0152] Specifically, a deep peak load distribution model is built according to the load forecast results, and the distribution results obtained by solving the model include:

[0153] The dynamic allocation parameters of power supply are calibrated according to the prediction results, and the dynamic allocation and power generation balance on the power supply day are preset as the target;

[0154] Based on the naive Bayes method, a combination method is used to express the objective function of the dynamic allocation of power supply on a daily basis;

[0155] On the basis of the objective function, a boundary function is set in combination with the power generation balance target, and the deep peak load distribution model is obtained by summing the objective function and the boundary function.

[0156] The required power generation is input into the allocation model to solve the output allocation results of photovoltaic and diesel complementary power supply during each peak load time.

[0157] The distribution verification module 304 is used to verify the accuracy of load distribution during complementary power supply and optimize the distribution model according to the verification results.

[0158] Specifically, verifying the accuracy of load distribution in complementary power supply and optimizing the distribution model according to the verification results includes the following steps:

[0159] Collect actual load distribution plans and actual operation data, including photovoltaic output, diesel generator output, load curve, etc., and calculate or measure the supply and demand balance in each period based on the actual operation data to determine whether there is excess or shortage of electricity.

[0160] Compare the actual supply and demand balance with the calculated results of the load distribution plan and analyze the deviation. If the deviation is too large, calculate the accuracy of the load distribution to determine whether it meets the requirements and analyze the reasons for the imbalance between supply and demand, which may be caused by load forecast deviation or inadequate distribution algorithm.

[0161] Improve the load forecasting model, retrain with more historical data, improve forecasting accuracy, adjust the distribution algorithm, modify algorithm parameters or optimize algorithm process, repeat verification until the load distribution achieves satisfactory accuracy, continue to collect operation data, monitor the long-term stability of load distribution accuracy, and make online adjustments when necessary to achieve closed-loop control and dynamically optimize the load distribution plan.

[0162] Specifically, through verification and optimization, the load distribution plan can be continuously improved to ensure the economy and stability of complementary power supply.

[0163] In summary, by improving the deep belief network to predict the complementary power supply load of photovoltaic and diesel, building a deep peak load distribution model and solving the distribution results, it can provide a quantitative reference for the operation of microgrids, guide the actual dispatching and control, realize the automatic and efficient configuration of microgrid resources, and improve the operation management level, which has positive significance for the operation and management of microgrids.

[0164] The monitoring and evaluation unit 4 is used to intelligently monitor the real-time operating status of the independent microgrid and evaluate the quality of power supply.

[0165] In this embodiment, intelligently monitoring the real-time operating status of an independent microgrid and evaluating the quality of power supply includes the following steps:

[0166] Sensors are deployed at key nodes of the microgrid to collect real-time data such as voltage, current, frequency, etc., as well as real-time power generation data of various distributed power sources such as photovoltaic, wind power, and diesel power generation.

[0167] Collect power consumption data on the load side, monitor load changes, use the communication network to aggregate the data of each node into the database of the monitoring center, develop a data analysis module, and calculate power quality evaluation indicators such as voltage quality, current balance, harmonic content, etc.

[0168] The state assessment algorithm is used to comprehensively consider various indicators to judge the operating state of the microgrid. When the assessment results show that the power supply quality has declined, an early warning or alarm is issued. The machine learning algorithm is used to train the assessment model to realize automatic assessment of the operating state of the microgrid.

[0169] Study the impact of different operating modes on power quality, give power supply optimization suggestions, link with the energy management system, and perform power supply optimization control based on the evaluation results.

[0170] Through intelligent monitoring and evaluation, real-time monitoring and continuous optimization of microgrid power supply quality can be achieved.

[0171] In summary, with the help of the above-mentioned technical scheme of the present invention, the present invention first estimates and analyzes the complementary power supply data of photovoltaic and diesel power supply in the microgrid, eliminates clutter interference by using the minimum mean difference method, and establishes a photovoltaic-diesel-load prediction model for deep peak-shaving optimization, thereby realizing the effective configuration and coordination of microgrid power supply resources and improving the overall power supply efficiency. By identifying power supply quality problems through data analysis, the equipment is adjusted and controlled in a targeted manner to improve the power supply quality. Not only can clutter interference be eliminated to effectively improve the power supply quality and reliability, but a prediction model can also be established to predict the supply and demand relationship in advance, perform deep peak-shaving and optimize load distribution, and reduce costs. Through data-driven predictive control means, efficient configuration and optimization of microgrid resources are realized, the overall power supply quality and operating efficiency are effectively improved, and the requirements of sustainable development are met. The present invention is based on establishing a simulation model by using signal data, determining the clutter signals generated during photovoltaic and diesel power generation, and performing classification identification and feature extraction, then constructing a clutter eliminator model, using a filter to estimate the variance of the clutter signal, using a minimum mean difference method to filter the variance to achieve clutter interference elimination, and finally verifying whether the clutter interference is successfully eliminated. The present invention can identify various types of clutter generated at different stages of photovoltaic and diesel power generation, which is conducive to targeted design of filters to eliminate different types of clutter, identifying and filtering clutter at the digital signal processing level instead of by adding additional equipment, which is more flexible and efficient, and adopts variance estimation and minimum mean difference filtering methods, which can effectively reduce the power of clutter, remove clutter interference to the maximum extent, improve the quality and reliability of photovoltaic and diesel power generation, and create conditions for stable power supply of microgrids. The present invention improves the deep belief network to predict the photovoltaic and diesel complementary power supply load, builds a deep peak-shaving load distribution model and solves the distribution results, which can better learn and express the photovoltaic-diesel complementary operation law and improve the accuracy of load prediction. Based on the load prediction results, the deep peak-shaving model can realize the efficient matching configuration of supply and demand resources, reduce energy consumption, provide quantitative reference for microgrid operation, guide actual dispatching and control, realize the automatic and efficient configuration of microgrid resources, improve the operation management level, and have positive significance for the operation and management of microgrid.

[0172] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A photovoltaic and diesel complementary power supply system in an independent microgrid, characterized in that: The photovoltaic and diesel complementary power supply system in the independent microgrid comprises a microgrid control unit (1), an interference elimination unit (2), a load optimization distribution unit (3) and a monitoring and evaluation unit (4); Wherein, the microgrid control unit (1) is connected to the interference elimination unit (2), the interference elimination unit (2) is connected to the load optimization allocation unit (3), and the load optimization allocation unit (3) is connected to the monitoring and evaluation unit (4); The microgrid control unit (1) is used to control photovoltaic and diesel power supply equipment in the microgrid, and to plan and adjust the strategy and status of the microgrid operation; The interference elimination unit (2) is used to estimate and analyze the electric energy data generated during complementary power supply, and eliminate clutter interference using a minimum mean difference method; The load optimization distribution unit (3) is used to establish a prediction model through the complementarity of photovoltaic and diesel power supply to optimize load distribution under deep peak regulation; The monitoring and evaluation unit (4) is used to intelligently monitor the real-time operating status of the independent microgrid and evaluate the quality of power supply; The interference elimination unit (2) comprises a data acquisition processing module (201), a clutter determination module (202), an estimation model construction module (203), a clutter elimination module (204) and a clutter verification module (205); The data acquisition and processing module (201) is connected to the clutter determination module (202), the clutter determination module (202) is connected to the estimation model construction module (203), the estimation model construction module (203) is connected to the clutter elimination module (204), and the clutter elimination module (204) is connected to the clutter verification module (205); The data acquisition and processing module (201) is used to collect electric energy data when photovoltaic and diesel are complementary to each other and perform preprocessing operations on the electric energy data, wherein the electric energy data includes the current and voltage when photovoltaic and diesel are powered, the voltage waveform of the connection point, the input and output current of the inverter when photovoltaic is powered, and the mechanical noise of the internal combustion engine when diesel is powered; The clutter determination module (202) is used to establish a simulation model using signal data to determine clutter signals generated during photovoltaic and diesel power supply, and to classify and extract features of the clutter signals; The estimation model building module (203) is used to build a clutter eliminator model using the clutter signal and formulate a filter to estimate the variance of the clutter signal; The clutter elimination module (204) is used to eliminate clutter interference by filtering the variance using the minimum mean difference; The clutter verification module (205) is used to verify whether there is clutter interference in photovoltaic and diesel power supply after clutter elimination; The method of using signal data to establish a simulation model to determine the clutter signals generated by photovoltaic and diesel power supply, and classifying and extracting features of the clutter signals includes: Based on the signal data, the input and output currents of the inverter when powered by photovoltaic power and the mechanical noise of the internal combustion engine when powered by diesel power are respectively equivalent to obtain a signal source, wherein the signal source includes a current source and a noise source; Combining the vibration, fluctuation, and phase difference mechanism during power supply with the Euler fusion algorithm, a simulation model is constructed, and the process of photovoltaic and diesel complementary power supply is simulated to obtain the phase difference; Based on the phase difference, a balance equation is used to calculate the clutter signal in different time periods, and based on the probability distribution equation, the composition rate of each part of the clutter signal is calculated, wherein the clutter signal includes voltage fluctuation and harmonics, load change, voltage fluctuation deviation and voltage flicker; Clutter signals are classified based on composition rate, and features of each group of classified data are extracted by combining genetic extraction algorithm; The simulation model is constructed by combining the vibration, fluctuation, phase difference mechanism during power supply with the Euler fusion algorithm, and the simulation process of photovoltaic and diesel complementary power supply is simulated, including: The simulation model is established by coupling photovoltaic and diesel using discrete dynamics and power transmission principles. Input the electric energy data into the simulation model to preset the vibration, fluctuation and phase difference data within different power supply durations; The fluctuations and vibrations generated by the simulation are taken as the continuous phase, the phase difference is taken as the discrete phase, and the discrete phase is tracked using the Euler fusion algorithm to determine the phase difference generated when the power supply is calculated based on the duration; The method of constructing a clutter eliminator model by using the clutter signal and formulating a filter to estimate the variance of the clutter signal includes: Construct a clutter model based on the extracted clutter signal features, design different clutter eliminators based on the clutter model, and establish the elimination capability index and threshold rules when the clutter eliminator is used; A clutter eliminator model is established based on the elimination capability index, and the clutter signal is input into the clutter eliminator model to generate adversarial data, and the adversarial data is normalized and the covariance matrix is ​​solved; Wavelet coefficients are used to solve the index weights of adversarial data, and the weighted sum method is used to aggregate the indexes, which are used as explicit variables of the clutter eliminator model. A clutter eliminator program is written according to the clutter eliminator model, and a maximum estimation method is used to perform calculations to obtain an estimated value of clutter in the clutter eliminator model; Determine whether the estimated value meets the threshold rule. If it is within the threshold range, it is retained; if it is not within the threshold range and does not meet the rule, it is discarded; Based on the estimated values, the quantitative relationship model between the elimination ability index and the manifest variable is obtained, and the elimination ability is evaluated; According to the evaluation results, a wavelet filter for clutter interference is constructed, and the wavelet filter is applied to the clutter signal for filtering to obtain an electric energy signal, and the variance of the electric energy signal is estimated; The method of using the minimum mean error variance to filter and eliminate clutter interference includes: Determine the window size of the waveform mean and variance of the electric energy signal, and use the minimum mean error method to filter the variance; The distribution characteristics of the preset electric energy signal are calculated to obtain the variance compactness information; Based on the variance compactness information, candidate initial clusters that meet the clutter interference elimination conditions are screened out, and the candidate initial clusters are processed and screened to obtain variances that meet the elimination conditions; The wavelet filter that can eliminate clutter interference is obtained by inverse analysis based on the variance results and is used for photovoltaic and diesel complementary power supply to ensure no clutter interference.

2. The photovoltaic and diesel complementary power supply system in an independent microgrid according to claim 1 is characterized in that: The calculation formula of the phase difference is: ; In the formula, W Indicates the phase difference value; represents a constant; e Indicates the current generated by diesel powered equipment; t represents the simulation time point; Indicates the current frequency value of diesel powered equipment; a Indicates the current value generated by photovoltaic power supply equipment; l Indicates the power supply duration.

3. The photovoltaic and diesel complementary power supply system in an independent microgrid according to claim 1 is characterized in that: The load optimization distribution unit (3) comprises a data collection and processing module (301), a load prediction module (302), a deep peak regulation module (303) and a distribution verification module (304); The data collection and processing module (301) is connected to the load prediction module (302), the load prediction module (302) is connected to the deep peak-shaving module (303), and the deep peak-shaving module (303) is connected to the allocation verification module (304); The data collection and processing module (301) is used to collect historical operating data of photovoltaic and diesel equipment and pre-process the operating data, wherein the operating data includes power output, load demand, fuel consumption and weather conditions on power supply days; The load prediction module (302) is used to predict the load of complementary power supply based on the operating principle of photovoltaic and diesel complementary power generation and select operating data using an improved deep belief network; The deep peak-shaving module (303) is used to build a deep peak-shaving load distribution model according to the load forecast result, and solve it to obtain the distribution result; The distribution verification module (304) is used to verify the accuracy of load distribution during complementary power supply, and optimize the distribution model according to the verification result.

4. The photovoltaic and diesel complementary power supply system in an independent microgrid according to claim 3 is characterized in that: The load forecasting of complementary power supply based on the operating principle of photovoltaic and diesel complementary power generation and selecting operating data using an improved deep belief network includes: According to the principle of photovoltaic power generation, the weather conditions on the power supply day are selected as the prediction data affecting the photovoltaic power supply, and the fuel consumption is selected as the prediction data affecting the diesel power supply; The two sets of prediction data are used as the input values ​​of the prediction model, and the improved deep belief network algorithm is used to generate the visible layer and the hidden layer; Using multi-layer restricted Boltzmann machine stacking, the load demand is used as the model output value to generate the power supply load forecasting mechanism input to the visual layer; The deep features of the input value are extracted and passed to the hidden layer to determine the hidden layer activation function, and the input value is extracted from the activation function to complete the update iteration to obtain the load forecasting model to achieve photovoltaic and diesel complementary power supply in line with the prediction.

5. The photovoltaic and diesel complementary power supply system in an independent microgrid according to claim 4 is characterized in that: The deep peak load distribution model is built according to the load forecast result, and the distribution result is obtained by solving the model. The dynamic allocation parameters of power supply are calibrated according to the prediction results, and the dynamic allocation and power generation balance on the power supply day are preset as the target; Based on the naive Bayes method, a combination method is used to express the objective function of the dynamic allocation of power supply on a daily basis; On the basis of the objective function, a boundary function is set in combination with the power generation balance target, and the deep peak load distribution model is obtained by summing the objective function and the boundary function. The required power generation is input into the allocation model to solve the output allocation results of photovoltaic and diesel complementary power supply during each peak load time.