Multi-time scale response analysis method and device based on random characteristics of flexible resources

Abstract

The invention discloses a multi-time scale response analysis method and device based on flexible resource random characteristics. The method comprises the following steps: acquiring flexible resource random response characteristics; establishing a flexible resource multi-time scale response model based on flexible resource random response characteristics; based on the flexible resource multi-time scale response model, a random response surface method is used to calculate a power grid electric energy quality index after the power grid node accesses the flexible resource; and analyzing the influence of the accessed flexible resources on the power quality of the power grid based on the calculated power quality index of the power grid, and optimizing a scheduling strategy of the flexible resources. The method optimizes the efficiency of the flexible resources participating in the power quality adjustment, and provides support for the precise regulation and control of the flexible resources and the safe and stable operation of a power grid.

Description

Technical Field

[0001] The present invention relates to a multi-time-scale response analysis method and device based on the random characteristics of flexible resources, belonging to the technical field of new energy regulation. Background Art

[0002] In the context of a new power system, the large-scale access of flexible resources such as distributed photovoltaics, energy storage, and electric vehicles has brought new challenges and opportunities to power grid operation. The adjustment methods of traditional power system equipment are relatively fixed and difficult to effectively cope with the impacts brought by the randomness and dynamics of distributed resources. However, the reactive power regulation ability of flexible resources, the charge and discharge regulation ability of energy storage devices, and the controllable response of residential loads provide new ideas for improving power quality problems.

[0003] The output of distributed photovoltaics has obvious randomness and intermittency, and its reactive power regulation function plays a key role in coping with voltage fluctuations and over-limit problems; the energy storage system can balance the output fluctuations of photovoltaics in terms of time and power through charge and discharge; electric vehicles and residential loads can respond to grid demands through charge and discharge or load transfer. Regarding the random characteristics and response behaviors of flexible resources, which show different performances on multi-time scales, the fast response ability is reflected on the short time scale, dynamic adjustment is required on the medium time scale, and the trend-based dispatching characteristics need to be analyzed on the long time scale.

[0004] Existing technologies mostly model the response characteristics of flexible resources on a single time scale and lack a comprehensive description of their random response behaviors on multi-time scales. How to accurately model and efficiently analyze their random response characteristics, and then provide support for the precise dispatching of flexible resources and the optimization of power quality, is an urgent problem to be solved currently. Summary of the Invention

[0005] The purpose of the present invention is to propose a multi-time-scale response analysis method and device based on the random characteristics of flexible resources, establish a multi-time-scale response model of flexible resources, calculate the power quality indexes of the power grid after accessing flexible resources at grid nodes by using the random response surface method, analyze the impact of the accessed flexible resources on the power quality of the power grid, and optimize the dispatching strategy of flexible resources.

[0006] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0007] In the first aspect, the present invention provides a multi-time-scale response analysis method based on the random characteristics of flexible resources, including:

[0008] Obtain the random response characteristics of flexible resources;

[0009] Based on the random response characteristics of flexible resources, establish a multi-time-scale response model of flexible resources;

[0010] Based on the flexible resource multi-time scale response model, the random response surface method is used to calculate the power quality index of the power grid after the flexible resources are connected to the grid nodes;

[0011] Based on the calculated power quality index of the power grid, analyze the impact of the connected flexible resources on the power quality of the power grid, which is used to optimize the scheduling strategy of the flexible resources.

[0012] Preferably, the flexible resources include distributed generation resources, energy storage systems and controllable loads; the distributed generation resources include distributed photovoltaic and wind turbines, and the energy storage systems include energy storage devices and electric vehicles;

[0013] The acquisition of the random response characteristics of flexible resources includes:

[0014] The random response characteristics of the distributed photovoltaic are shown as:

[0015] ,

[0016] Where: is the output power of the distributed photovoltaic unit at time , is the light curtailment rate of the output of the distributed photovoltaic unit at time , is the rated power of the distributed photovoltaic unit corrected according to the meteorological parameters , is the light intensity of the distributed photovoltaic unit , is the temperature of the distributed photovoltaic unit , is the prediction deviation;

[0017] At the same time, the reactive power constraint of the distributed photovoltaic needs to be satisfied;

[0018] The random response characteristics of the energy storage system are shown as:

[0019] ,

[0020] Where, is the SOC value of the energy storage system at time , is the initial SOC value of the energy storage system , and are the charge and discharge power of the energy storage system at time , is the charge and discharge efficiency, is the time interval, For an energy storage system Maximum capacity;

[0021] Meanwhile, it is necessary to satisfy the power constraint and SOC constraint of the energy storage system;

[0022] The random response characteristics of the controllable load are manifested as:

[0023] ,

[0024] Among them, is the regulating power of the controllable load at time , is the maximum adjustable power of the controllable load at time , is the load power after prediction deviation correction, is the predicted power of the controllable load at time , is the prediction error;

[0025] Meanwhile, it is necessary to satisfy the power constraint of the controllable load.

[0026] Preferably, based on the random response characteristics of flexible resources, a multi-time scale response model of flexible resources is established, including:

[0027] For the random response characteristics of distributed photovoltaic, a short-time scale dynamic response model is constructed by using the time series analysis method, and the short-time scale refers to the time scale of minutes;

[0028] For the charge and discharge random response characteristics of the energy storage system, a medium-time scale dynamic response model is established by using the state space model or Markov chain, and the medium-time scale refers to the time scale of 15 minutes;

[0029] For the random response characteristics of the controllable load, a long-time scale dynamic response model is established by using the probability distribution fitting method, and the long-time scale refers to the time scale of hours.

[0030] Preferably, based on the multi-time scale response model of flexible resources, the power quality index of the power grid after accessing flexible resources is calculated by using the random response surface method, including:

[0031] Random variables are selected for each flexible resource as the input of the multi-time scale response model of flexible resources;

[0032] Orthogonal polynomials are selected for each flexible resource. Based on the selected orthogonal polynomials, the multi-time scale response model of flexible resources is converted into the form of orthogonal polynomial expansion by using the random response surface method;

[0033] Obtain a set of sample points of the input random variable, perform polynomial fitting, and obtain the fitting coefficients of the polynomial;

[0034] Based on the expanded form of the orthogonal polynomial after fitting, calculate the power quality indexes of the power grid after the flexible resources are connected to the power grid nodes; the power quality indexes of the power grid include voltage, frequency, and reactive power.

[0035] Preferably, selecting random variables for each flexible resource includes:

[0036] The input random variables of distributed generation resources are light intensity, temperature, and prediction deviation, and they follow a Beta distribution;

[0037] The input random variables of energy storage devices are the initial state of charge and charge and discharge power, and they follow a normal distribution; the input random variable of electric vehicles is the connection time, and it follows a normal distribution;

[0038] The input random variables of controllable loads are the maximum power that the controllable load can adjust and the prediction error, and they follow a normal distribution or a uniform distribution.

[0039] Preferably, selecting orthogonal polynomials for each flexible resource includes:

[0040] Select Jacobi polynomials for distributed generation resources;

[0041] Select Hermite polynomials for energy storage devices and electric vehicles;

[0042] If the input random variable of the controllable load follows a normal distribution, select Hermite polynomials; if the input random variable follows a uniform distribution, select Legendre polynomials.

[0043] Preferably, the obtaining of the set of sample points of the input random variable includes:

[0044] Use the Latin hypercube sampling or Monte Carlo sampling method to generate a set of sample points of the input random variable;

[0045] The performing of polynomial fitting to obtain the fitting coefficients of the polynomial includes: using the least squares method to perform polynomial fitting to obtain the fitting coefficients of the polynomial.

[0046] In a second aspect, the present invention also provides a multi-time scale response analysis device based on the random characteristics of flexible resources, which is used to implement the multi-time scale response analysis method based on the random characteristics of flexible resources. The device includes:

[0047] A data collection module, which is used to obtain the random response characteristics of flexible resources;

[0048] A modeling module, configured to establish a multi-time scale response model of flexible resources based on the random response characteristics of flexible resources;

[0049] A calculation module, configured to calculate the power grid power quality index after the flexible resources are connected to the power grid nodes by using the random response surface method based on the multi-time scale response model of the flexible resources;

[0050] An analysis module, configured to analyze the impact of the connected flexible resources on the power grid power quality based on the calculated power grid power quality index, and to optimize the scheduling strategy of the flexible resources.

[0051] In a third aspect, the present invention further provides a computer-readable storage medium storing one or more programs, where the one or more programs include instructions that, when executed by a computing device, cause the computing device to execute any of the above multi-time scale response analysis methods based on the random characteristics of flexible resources.

[0052] In a fourth aspect, the present invention further provides a computing device, including one or more processors, a memory, and one or more programs, where the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for executing any of the above multi-time scale response analysis methods based on the random characteristics of flexible resources.

[0053] The beneficial effects of the present invention are as follows:

[0054] The present invention proposes a multi-time scale response analysis method based on the random characteristics of flexible resources, establishes a multi-time scale response model of flexible resources, calculates the power grid power quality index after the flexible resources are connected to the power grid nodes by using the random response surface method, analyzes the impact of the connected flexible resources on the power grid power quality, and optimizes the scheduling strategy of the flexible resources. The present invention provides support for accurately regulating flexible resources and supporting the safe and stable operation of the power grid by optimizing the efficiency of flexible resources participating in power quality regulation. Description of the Drawings

[0055] Figure 1 It is a schematic flowchart of the multi-time scale response analysis method based on the random characteristics of flexible resources provided by an embodiment of the present invention;

[0056] Figure 2 It is a topology diagram of a distribution network power grid provided in an embodiment of the present invention. Detailed Embodiments

[0057] To make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below in conjunction with the embodiments and the drawings. Herein, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but do not limit the present invention.

[0058] Here, it should also be noted that in order to avoid obscuring the present invention with unnecessary details, only the structures and / or processing steps closely related to the solution according to the present invention are shown in the drawings, while other details less relevant to the present invention are omitted.

[0059] It should be emphasized that the term "comprising / including" when used herein refers to the presence of features, elements, steps or components, but does not exclude the presence or addition of one or more other features, elements, steps or components.

[0060] It should be emphasized here that the step marks mentioned hereinafter do not limit the order of the steps. Instead, it should be understood that the steps can be executed in the order mentioned in the embodiments, or different from the order in the embodiments, or several steps can be executed simultaneously.

[0061] The embodiment of the present invention provides a multi-time scale characterization and analysis method based on the random response of flexible resources, and the method includes the following steps:

[0062] S1. Obtain the random response characteristics of flexible resources;

[0063] It should be noted that flexible resources include, in the context of a new power system: distributed generation resources represented by distributed photovoltaics and wind turbines; demand-side loads including adjustable loads and uninterruptible loads; energy storage devices mainly including energy storage and electric vehicles. The output of distributed generation resources is affected by external factors, making it difficult to accurately characterize its uncertain parameters and solve them efficiently; the electricity consumption characteristics of demand-side loads are flexible and can be quickly responded to by factors such as supply-demand balance and price signals in the short term, and include load start-up status, operation progress, action constraints of adjustable equipment, etc.; energy storage devices present non-linear time-varying electricity consumption characteristics, and their energy interaction process includes charge and discharge power, indicator functions, charge and discharge efficiency, etc.

[0064] The random response is defined as: the rapid response ability of flexible resources to grid voltage fluctuations and over-limit behaviors under uncertain conditions (such as the randomness of distributed photovoltaic output, the time difference of electric vehicle charging and discharging behaviors, etc.). It does not fix a certain mode, but is an efficient response from a global analysis perspective. It includes response characteristics of real-time (minute level), dynamic (15-minute level) and trend (hour level).

[0065] In this embodiment, the random response model of flexible resources includes:

[0066] (1) Random response characteristics of distributed photovoltaics:

[0067] The random response of distributed photovoltaics is mainly affected by meteorological conditions such as light intensity and temperature, and its output power has the characteristics of intermittency and randomness.

[0068] The output power model is as follows:

[0069] ,

[0070] where: is the output power of the distributed photovoltaic unit at time is the curtailment rate of the output of the distributed photovoltaic unit at time is the rated power of the distributed photovoltaic unit after being corrected according to meteorological parameters, is the light intensity of the distributed photovoltaic unit at time is the temperature of the distributed photovoltaic unit at time is the prediction deviation.

[0071] It should be noted that establishing the performance curve of the photovoltaic module and correcting it according to the actual light intensity and temperature measurement data to obtain the corrected rated power belongs to the common knowledge in this technical field.

[0072] The random response characteristics of the distributed photovoltaic need to satisfy the reactive power constraint:

[0073] ,

[0074] where is the actual reactive power that can be output by the inverter of the distributed photovoltaic unit at time is the rated maximum reactive power output of the inverter of the distributed photovoltaic unit at time. This constraint ensures that the reactive power output of the photovoltaic inverter does not exceed its rated capacity.

[0075] (2) Random response characteristics of the energy storage system:

[0076] The random response of the energy storage system is mainly affected by constraints such as the charge-discharge state and SOC (state of charge). The randomness of the energy storage device is reflected in: the random fluctuation of the initial SOC state, the dynamic changes of the charge-discharge efficiency and the random power demand.

[0077] SOC dynamic model:

[0078] ,

[0079] where: and are the charge and discharge powers of the energy storage system at time Charge and discharge efficiency; Time interval; is the SOC value of the energy storage system at a certain moment, indicating the current energy state of the energy storage device, is the energy storage system initial SOC value, is the energy storage system maximum capacity.

[0080] It is necessary to meet the power constraint:

[0081] ,

[0082] SOC constraint:

[0083] .

[0084] is the energy storage system maximum charge and discharge power, , is the energy storage system the maximum and minimum SOC allowed for charge and discharge.

[0085] (3) Random response characteristics of controllable load:

[0086] The random response of the controllable load is reflected in the load reduction and transfer capabilities, as well as the uncertainty of user response.

[0087] Controllable load regulation power model:

[0088] ,

[0089] Where: is the regulation power of the controllable load at a certain moment , is the maximum adjustable power of the controllable load at a certain moment , is the load power after prediction deviation correction, is the predicted power of the controllable load at a certain moment , is the prediction error. It should be noted that analyzing the user load response characteristics and correcting the predicted power by establishing a correction formula belong to the common knowledge in this technical field.

[0090] It is necessary to meet the power constraint:

[0091] ,

[0092] Among them, , are time-controllable loads reactive power and maximum output reactive power.

[0093] S2. Based on the stochastic response characteristics of flexible resources, establish a multi-time scale response model as follows:

[0094] Short time scale (minute level): For the fast reactive power regulation capabilities of distributed photovoltaic and energy storage, construct a short time scale dynamic response model to describe their fast regulation behavior for voltage fluctuations. Use time series analysis methods (such as ARIMA) to extract second-level response characteristics.

[0095] Medium time scale (15-minute level): Describe the charge and discharge dynamic adjustment behavior of energy storage devices and the role of electric vehicles in regulation. Use state space models or Markov chains to establish medium time scale dynamic response models.

[0096] Long time scale (hour level): For behaviors such as residential load transfer and long-term energy storage scheduling, extract their trend characteristics and use probability distribution fitting methods to establish long time scale dynamic response models.

[0097] S3. Based on the multi-time scale response model, use the Stochastic Response Surface Method (SRSM) to analyze and optimize the stochastic response characteristics of flexible resources as follows:

[0098] Model the input variables of flexible resources (such as photovoltaic irradiance, electric vehicle access time, etc.) as random variables, and combine orthogonal polynomials (such as Hermite, Laguerre, Jacobi, etc.) to construct a stochastic response surface. The output variables are power grid power quality indicators (such as voltage, frequency, power factor, etc.). Use the stochastic response surface method to quickly calculate the statistical characteristics (such as mean, variance) of the output and sensitivity analysis.

[0099] S31. Select random variables for each flexible resource and determine the response variables,

[0100] In this embodiment, the random variables are selected as:

[0101] (1) Input random variables:

[0102] Distributed photovoltaic: light intensity , temperature , prediction deviation ;

[0103] Energy storage system: initial SOC value , charge and discharge power , ;

[0104] Controllable load: The maximum power that the controllable load can be adjusted , prediction error .

[0105] (2) Response variable:

[0106] Power quality index: Voltage , frequency , reactive power ;

[0107] The power regulation ability of the resource.

[0108] S32. Select orthogonal polynomials for each flexible resource,

[0109] In this embodiment, according to the characteristics of the flexible resources, suitable orthogonal polynomials are selected for each type of resource as follows:

[0110] (1) Distributed photovoltaic and wind turbines:

[0111] Characteristics: The output power is mainly affected by light intensity, wind speed, etc., and usually shows a Beta distribution.

[0112] Selection: Jacobi polynomial (suitable for Beta distribution).

[0113] (2) Charge and discharge of energy storage system:

[0114] Characteristics: The randomness of the charge and discharge behavior is mainly determined by the equipment status and demand fluctuations, and is close to a normal distribution.

[0115] Selection: Hermite polynomial (suitable for normal distribution).

[0116] (3) Electric vehicle charge and discharge behavior:

[0117] Characteristics: The access time and charging demand have obvious concentration (for example, accessing during peak hours), showing a normal distribution.

[0118] Selection: Hermite polynomial.

[0119] (4) Controllable load transfer and reduction:

[0120] Characteristics: The response of the controllable load is related to user behavior and has strong randomness, but can be modeled as a specific distribution (such as normal distribution or uniform distribution).

[0121] Selection: Adjusted to Hermite polynomial or Legendre polynomial according to the load distribution characteristics.

[0122] S33. Based on the selected orthogonal polynomials, through the stochastic response surface method, the multi-time scale response model of flexible resources is transformed into the form of orthogonal polynomial expansion, expressed as:

[0123] ,

[0124] where: is the output power grid power quality index (such as voltage, frequency, power factor, etc.), is the input random variable, is the orthogonal polynomial basis function, is the orthogonal polynomial coefficient, determined by fitting the sample points.

[0125] The form of orthogonal polynomial expansion of each flexible resource is as follows:

[0126] (1) Distributed photovoltaic and wind turbines (modeled by Jacobi polynomials)

[0127] Let the input random variable be , assuming it follows a Beta distribution; the distribution parameter is the shape parameter of the Beta distribution.

[0128] The orthogonal polynomial selected is the Jacobi polynomial:

[0129] ,

[0130] where , is the polynomial order.

[0131] The orthogonal polynomial expansion form of the multi-time scale response model of distributed photovoltaic and wind turbines is:

[0132] ,

[0133] is the orthogonal polynomial expansion of the distributed photovoltaic output, is the fitting coefficient, which needs to be fitted through sampling data, is the highest order of the orthogonal polynomial expansion, or the truncation order of the expansion.

[0134] (2) Energy storage device charging and discharging (modeled by Hermite polynomials)

[0135] Let the input random variable be , assuming it follows a normal distribution.

[0136] The orthogonal polynomial selected is the Hermite polynomial:

[0137] ,

[0138] The orthogonal polynomial expansion form of the multi-time scale response model of the energy storage system is as follows:

[0139] ,

[0140] is the orthogonal polynomial expansion of the charge and discharge power of the energy storage system, is the fitting coefficient, which needs to be fitted through sampling data.

[0141] (3) Electric vehicle charging and discharging (Hermite polynomial modeling)

[0142] Let the input random variable be the access time, assumed to be normally distributed.

[0143] The orthogonal polynomial selects the Hermite polynomial:

[0144] ,

[0145] The orthogonal polynomial expansion form of the multi-time scale response model of the electric vehicle is as follows:

[0146] ,

[0147] is the orthogonal polynomial expansion of the charge and discharge power of the electric vehicle, is the fitting coefficient, which needs to be fitted through sampling data.

[0148] (4) Controllable load (Hermite or Legendre polynomial modeling)

[0149] Let the input random variable be , regarded as normally distributed or uniformly distributed.

[0150] The orthogonal polynomial is selected according to the characteristics of the research object or weighted processing is carried out for different time periods:

[0151] Table 1 Comparison of the characteristics of Hermite and Legendre polynomials

[0152] Characteristic Hermite polynomial Legendre polynomial Random variable distribution Normal distribution Uniform distribution User load response characteristics Concentrated during peak periods with obvious symmetry Uniformly distributed within a certain time period or load range Application scenario Centralized response of user behavior to time or load reduction Wide distribution range of load response, evenly dispersed behavior

[0153] If the orthogonal polynomial selects the Legendre polynomial:

[0154]

[0155] S34. Obtain the sample point set of the input random variable, perform polynomial fitting, and obtain the fitting coefficient,

[0156] In this embodiment, the Latin Hypercube Sampling (LHS) or Monte Carlo sampling method is used to generate a set of sample points of the input random variables;

[0157] The corresponding output variables are calculated through the flexible resource stochastic response model.

[0158] In this embodiment, the least squares method is used to obtain the fitting coefficients, expressed as:

[0159] ;

[0160] Finally, based on the orthogonal polynomial expansion form of the multi-time scale response model of each flexible resource, the response output of each flexible resource is calculated.

[0161] S35. Analyze and evaluate based on the output results of the response variables of each flexible resource, and optimize the scheduling strategy at different time scales.

[0162] In this embodiment, the response output characteristics of the flexible resources are analyzed and evaluated through the mean and variance as follows:

[0163] (1) Mean: It reflects the expected value of the system output variable (power quality index) and is an average description of the overall operating state of the system. The mean is used to evaluate the overall operating level of the power grid, such as the average contribution of the reactive power regulation of distributed photovoltaics to the system voltage.

[0164] ,

[0165] is the output response variable.

[0166] (2) Variance: It reflects the volatility or uncertainty degree of the output variable. It is used to evaluate the system stability. The smaller the value, the better the regulation effect of the flexible resource on the power grid.

[0167] .

[0168] Combined with the variance change rate, judge the sensitivity of the uncertainty of the random variable to the output index, that is, sensitivity analysis. Furthermore, evaluate the contribution of each input random variable to the output response, and determine the key factors for regulation and optimization.

[0169] In this embodiment, the Sobol index decomposition method is used for global sensitivity analysis. It should be noted that the sensitivity analysis methods include but are not limited to the Sobol index decomposition method, and the rest of the traditional variance decomposition methods should also be included in the scope of protection.

[0170] ,

[0171] is the first-order sensitivity of the variable point; is the output mean under given conditions at variable points. Variables with high sensitivity have a great impact on the output and should be optimized first. Variables with low sensitivity have a small impact on the output and their model complexity can be reduced.

[0172] Based on the results of stochastic response analysis, optimize the scheduling strategies of flexible resources at different time scales as follows:

[0173] Short time scale: Respond quickly to voltage fluctuations, adjust the reactive power output through distributed photovoltaic inverters, and the energy storage system discharges quickly to suppress the fluctuations.

[0174] Medium time scale: Dynamically adjust the load and energy storage status, optimize the charging and discharging time of electric vehicles, and coordinate the charging strategies of energy storage devices.

[0175] Long time scale: Balance the load and resource consumption, formulate a residential load transfer plan, and smooth the load curve.

[0176] To verify the effectiveness of the model and method proposed in this embodiment, taking an actual distribution area in Jiangsu as an example, set up a case for simulation analysis. This distribution area includes the access of distributed photovoltaics, energy storage systems, electric vehicles and controllable loads, and the operation of the distribution area faces typical voltage fluctuations and load peak-valley differences. Model and analyze the flexible resources through the multi-time scale response analysis method based on the stochastic characteristics of flexible resources proposed in this embodiment to verify the improvement effect on power quality.

[0177] As Figure 2 shown, the distributed photovoltaics are connected to node 4 (80 kW) and node 10 (80 kW) respectively, with a total capacity of 160 kW. In terms of stochastic characteristics, the photovoltaic output power follows a Beta distribution ; the energy storage system is connected to node 1, with a rated power of 80 kW and a capacity of 160 kWh. In terms of stochastic characteristics, the initial SOC follows a normal distribution with a mean of 50% and a variance of 10%; the electric vehicles are connected to node 8, and the total power of the charging piles is 21 kW. In terms of stochastic characteristics, the access time follows a normal distribution with a mean of 19:00 and a variance of 0.5 hours; the total controllable load is 30 kW; stochastic characteristics: the response characteristics follow a uniform distribution of [5, 15] kW.

[0178] The grid constraints refer to the national standard low-voltage distribution area voltage standard (GB / T 12325-2008), and the per-unit value is set to [0.95, 1.05]; the photovoltaic reactive power regulation range is 20% of the rated capacity. Use Latin hypercube sampling (LHS) to generate 1000 random input sample points, and sample random variables such as photovoltaic irradiance, energy storage SOC, electric vehicle access time, and load reduction.

[0179] Based on the above multi-time scale response analysis method based on the stochastic characteristics of flexible resources, conduct modeling analysis as follows:

[0180] (1)Construct a stochastic response model for flexible resources

[0181] Use Jacobi polynomials to describe the stochastic output of photovoltaic power; use Hermite polynomials to describe the charging and discharging behaviors of energy storage and electric vehicles; according to the uniform distribution characteristics, use Legendre polynomials to describe the transfer and reduction characteristics of controllable loads.

[0182] (2)Stochastic response surface fitting

[0183] Construct a stochastic response surface to fit the model coefficients by the least squares method , and calculate the node voltage fluctuation (output index).

[0184] (3)Sensitivity analysis

[0185] Calculate the Sobol index of the input random variables on voltage fluctuation to identify the factors that have the greatest impact on power quality.

[0186] (4)Dispatch optimization

[0187] The optimization strategies include: dynamically adjusting the reactive power output of photovoltaic power, the charging and discharging plans of energy storage, the access time of electric vehicles, and the transfer time of controllable loads.

[0188] The following Table 2 shows the comparison of the mean and variance of voltage fluctuations at the main nodes before and after optimization.

[0189] Table 2 Comparison of the mean and variance of voltage fluctuations at the main nodes before and after optimization

[0190] Node number Mean value before optimization (p.u.) Mean value after optimization (p.u.) Variance before optimization Variance after optimization Fluctuation reduction rate 1 1.052 1.035 0.0054 0.0029 46.3% 4 1.060 1.038 0.0060 0.0033 45.0% 8 1.057 1.034 0.0056 0.0030 46.4% 10 1.059 1.036 0.0058 0.0032 44.8%

[0191] Analysis shows that after optimization, the mean voltage of all nodes is stable within the allowable range, the variance of voltage fluctuation is greatly reduced, and the system stability is significantly improved.

[0192] The analysis results of the Sobol index of the input variables are shown in Table 3 below.

[0193] Table 3 Sensitivity analysis table

[0194]

[0195] Analysis shows that the photovoltaic output and the SOC of energy storage are the main factors affecting voltage fluctuation; the access behavior of electric vehicles is the second; the contribution of load reduction is small, and the optimization can be appropriately simplified.

[0196] Based on this, in terms of photovoltaic reactive power regulation, the curtailment rate can be controlled below 5%, and the reactive power output of the photovoltaic inverter can be dynamically adjusted. Energy storage is dynamically adjusted, discharging during peak hours and charging during valley hours, and preferentially adjusting the energy storage SOC to a reasonable range. The access of electric vehicles is optimized to adjust the access timing and reduce the concentration of access time. The controllable load cuts 10 kW of load during peak hours.

[0197] Based on the above analysis, this embodiment has adaptability for the following different actual scenarios.

[0198] Table 4 Summary of adaptability in different scenarios

[0199] Scenario Targeted problem Solution Scenario 1: High-penetration distributed photovoltaic access to the distribution network Voltage over-limit caused by photovoltaic output fluctuation Optimize the reactive power regulation strategy of photovoltaic using a stochastic response model Scenario 2: Large-scale access of electric vehicles to the distribution network Voltage fluctuation and overload caused by the uncertainty of charging behavior Optimize the timing scheduling of electric vehicle access based on a stochastic response surface Scenario 3: Flexible response of residential load Excessive peak-valley difference of load Formulate a load transfer strategy by combining the stochastic response characteristics of controllable loads

[0200] Based on the above inventive concept, another embodiment of the present invention provides a multi-time scale response analysis device based on the stochastic characteristics of flexible resources, for implementing the above multi-time scale response analysis method based on the stochastic characteristics of flexible resources. The device includes:

[0201] A data collection module for obtaining the stochastic response characteristics of flexible resources;

[0202] A modeling module for establishing a multi-time scale response model of flexible resources based on the stochastic response characteristics of flexible resources;

[0203] A calculation module for calculating the power quality index of the power grid after accessing flexible resources at the power grid nodes by using the stochastic response surface method based on the multi-time scale response model of flexible resources;

[0204] An analysis module for analyzing the impact of the accessed flexible resources on the power quality of the power grid based on the calculated power quality index of the power grid, and for optimizing the scheduling strategy of flexible resources.

[0205] It should be noted that the device embodiment corresponds to the above method embodiment, and the implementation manners of the above method embodiment are all applicable to the device embodiment and can achieve the same or similar technical effects, so they will not be elaborated here.

[0206] Based on the above inventive concept, another embodiment of the present invention provides a computer-readable storage medium storing one or more programs, where the one or more programs include instructions that, when executed by a computing device, cause the computing device to execute the multi-time scale response analysis method based on the stochastic characteristics of flexible resources as described above.

[0207] Based on the above inventive concept, another embodiment of the present invention provides a computing device, including one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for executing the multi-time scale response analysis method based on the random characteristics of flexible resources as described above.

[0208] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0209] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processors of the computer or other programmable data processing devices generate means for implementing the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0210] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means, and the instruction means implements the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0211] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0212] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: it is still possible to modify the specific implementation manners of the present invention or make equivalent replacements, and any modification or equivalent replacement that does not depart from the spirit and scope of the present invention shall be covered by the protection scope of the claims of the present invention.

Claims

1. A multi-time scale response analysis method based on the random characteristics of flexible resources, characterized in that: include: Obtaining the random response characteristics of flexible resources; Based on the random response characteristics of flexible resources, a flexible resource multi-time scale response model is established; Based on the flexible resource multi-time scale response model, the random response surface method is used to calculate the power quality index of the power grid after the power grid node is connected to the flexible resource; Based on the calculated power quality index of the power grid, the impact of the connected flexible resources on the power quality of the power grid is analyzed to optimize the scheduling strategy of the flexible resources.

2. According to claim 1, a multi-time scale response analysis method based on the random characteristics of flexible resources is characterized in that: The flexible resources include distributed power generation resources, energy storage systems and controllable loads; the distributed power generation resources include distributed photovoltaics and wind turbines, and the energy storage system includes energy storage equipment and electric vehicles; The random response characteristics of the distributed photovoltaic are as follows: ， in: for Distributed photovoltaic units at all times The output power, for Distributed photovoltaic units at all times The abandoned light rate of the output, For distributed photovoltaic units Rated power corrected according to meteorological parameters, For distributed photovoltaic units The light intensity, For distributed photovoltaic units The temperature, Forecast bias; At the same time, the reactive power constraints of distributed photovoltaics must be met.

3. The multi-time scale response analysis method based on the random characteristics of flexible resources according to claim 2 is characterized in that: The random response characteristics of the energy storage system are as follows: ， in, for Momentary energy storage system SOC value, For energy storage system Initial SOC value, and for Momentary energy storage system The charging and discharging power, Charge and discharge efficiency, time interval, For energy storage system Maximum capacity; At the same time, the power constraints and SOC constraints of the energy storage system must be met.

4. The multi-time scale response analysis method based on the random characteristics of flexible resources according to claim 2 is characterized in that: The random response characteristics of the controllable load are as follows: ， in, for Controllable load at all times Adjust the power, for Controllable load at all times Adjustable maximum power, is the load power after forecast deviation correction, for Controllable load at all times Predicted power, is the prediction error; At the same time, the power constraints of the controllable load must be met.

5. A multi-time scale response analysis method based on the random characteristics of flexible resources according to claim 1 or 2, characterized in that: The flexible resource multi-time scale response model is established based on the random response characteristics of the flexible resource, including: In view of the random response characteristics of distributed photovoltaics, a short-time scale dynamic response model is constructed using time series analysis method, where the short-time scale refers to the time scale of minutes; In view of the random response characteristics of charging and discharging of the energy storage system, a medium-time scale dynamic response model is established using a state space model or a Markov chain. The medium-time scale refers to a time scale of 15 minutes. According to the random response characteristics of controllable loads, a long-time scale dynamic response model is established by using a probability distribution fitting method. The long-time scale refers to a time scale of hours.

6. The multi-time scale response analysis method based on the random characteristics of flexible resources according to claim 5 is characterized in that: The method of calculating the power quality index of a power grid after a power grid node is connected to a flexible resource by using a random response surface method based on the flexible resource multi-time scale response model includes: Selecting random variables for each flexible resource as input to the flexible resource multi-time scale response model; An orthogonal polynomial is selected for each flexible resource, and based on the selected orthogonal polynomial, the multi-time scale response model of the flexible resource is converted into the form of orthogonal polynomial expansion through random response surface method; Obtain a sample point set of the input random variable, perform polynomial fitting, and obtain the fitting coefficients of the polynomial; Based on the expanded form of the fitted orthogonal polynomial, power quality indicators of the power grid after the power grid nodes are connected to the flexible resources are calculated; the power quality indicators of the power grid include voltage, frequency and reactive power.

7. The multi-time scale response analysis method based on the random characteristics of flexible resources according to claim 6 is characterized in that: The step of selecting a random variable for each flexible resource includes: The random variables of the distributed generation resource input are light intensity, temperature and prediction deviation, which follow Beta distribution; The input random variables of energy storage equipment are initial state of charge and charge / discharge power, which obey normal distribution; the input random variables of electric vehicles are access time, which obey normal distribution; The controllable load input random variables are the maximum power and prediction error that the controllable load can adjust, and they obey normal distribution or uniform distribution.

8. The multi-time scale response analysis method based on the random characteristics of flexible resources according to claim 7 is characterized in that: The step of selecting an orthogonal polynomial for each flexible resource includes: Jacobi polynomial for distributed generation resource selection; Energy storage devices and electric vehicles select Hermite polynomials; If the input random variable of controllable load follows normal distribution, choose Hermite polynomial; if the input random variable follows uniform distribution, choose Legendre polynomial.

9. The multi-time scale response analysis method based on the random characteristics of flexible resources according to claim 8 is characterized in that: The step of obtaining a sample point set of an input random variable includes: Generate a set of sample points of the input random variable using Latin hypercube sampling or Monte Carlo sampling method; The performing polynomial fitting to obtain the fitting coefficients of the polynomial includes: performing polynomial fitting using the least squares method to obtain the fitting coefficients of the polynomial.

10. A multi-time scale response analysis device based on the random characteristics of flexible resources, characterized in that: The device is used to implement the multi-time scale response analysis method based on the random characteristics of flexible resources according to any one of claims 1 to 9, comprising: Data collection module, used to obtain the random response characteristics of flexible resources; A modeling module is used to establish a multi-time scale response model of flexible resources based on the random response characteristics of flexible resources; A calculation module, used to calculate the power quality index of the power grid after the power grid node is connected to the flexible resource by using the random response surface method based on the flexible resource multi-time scale response model; The analysis module is used to analyze the impact of the connected flexible resources on the power quality of the power grid based on the calculated power quality index of the power grid, so as to optimize the scheduling strategy of the flexible resources.

11. A computer-readable storage medium storing one or more programs, characterized in that: The one or more programs include instructions, which, when executed by a computing device, enable the computing device to perform any one of the multi-time scale response analysis methods based on random characteristics of flexible resources according to claims 1 to 9.

12. A computing device, characterized in that: The method comprises one or more processors, a memory and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for executing any one of the multi-time scale response analysis methods based on the random characteristics of flexible resources according to claims 1 to 9.

Images