Quantitative evaluation method for adjustable capability of light, storage, charging and load resources in power distribution network district

Abstract

The invention discloses a quantitative evaluation method for the adjustable capability of light, storage, charging and load resources in a power distribution network area. The method comprises the following steps: S1, data acquisition: acquiring photovoltaic output, energy storage SOC, charging pile charging demand and load power consumption data through an intelligent terminal; s2, data preprocessing: carrying out normalization processing on the acquired data by adopting a minimum and maximum normalization method, a Zscore standardization method and a one-hot coding method; s3, feature modeling; S4, dynamic scene correction: establishing a state space and an action space through a Markov decision process, and constructing a reward function in combination with an electric price difference and equipment loss; and S5, comprehensive evaluation: calculating the comprehensive score of the adjustable capability of each resource by adopting an entropy weight TOPSIS method, and generating a transformer area adjustment potential report. According to the method, the adjustable capability of resources such as light, storage, charging and load can be accurately quantified, a multi-resource collaborative modeling framework is proposed for the first time, and the evaluation precision and practicability are remarkably improved in combination with a dynamic correction mechanism driven by real-time data.

Description

Technical Field

[0001] The present invention relates to the technical field of image digital processing, and in particular to a method for quantitatively evaluating the adjustable capabilities of light, storage, charging and load resources within a distribution network area. Background Art

[0002] With the rapid development of power systems, the number of distributed resources within distribution networks (such as photovoltaics, energy storage, charging stations, and adjustable loads) has steadily increased. The integration of these resources poses new challenges to the stable operation and dispatch of power grids. Distributed resources have varying regulation characteristics and are influenced by a variety of factors, including the environment, user behavior, and device status. Traditional power system operation and control methods are no longer able to meet the demands of modern distribution networks.

[0003] Existing power system operation and control technologies primarily focus on evaluating the regulation capabilities of a single resource, such as the adjustability of photovoltaic power generation or the charge and discharge capabilities of energy storage systems. However, the diverse range of distributed resources within distribution network substations, coupled with complex spatiotemporal coupling, makes single-resource evaluation methods incapable of fully reflecting the coordinated regulation potential of multiple resources within a substation.

[0004] The defects and shortcomings of the prior art are:

[0005] 1. Lack of multi-resource collaborative evaluation methods:

[0006] Existing technologies mostly evaluate the adjustable capacity of a single resource (such as photovoltaics, energy storage, charging piles or loads), and lack a quantitative assessment method for the coordinated regulation potential of multiple resources.

[0007] Since the regulation characteristics of different resources vary greatly, the evaluation method of a single resource cannot accurately reflect the overall regulation capacity of multiple resources within the substation area.

[0008] 2. Ignoring the spatiotemporal coupling relationship between resources:

[0009] Existing technologies often overlook the spatiotemporal coupling between resources such as photovoltaics, energy storage, charging stations, and loads. For example, photovoltaic power generation is affected by the intensity and duration of sunlight, the charge and discharge capacity of energy storage systems is limited by SOC (State of Charge), and charging demand at charging stations is influenced by user behavior. These factors interact in a complex manner.

[0010] Ignoring these coupling relationships will lead to inaccurate evaluation results and fail to provide effective decision support for power grid dispatching.

[0011] 3. Insufficient accuracy of static evaluation methods:

[0012] Existing technologies often use static evaluation methods, assuming that resource scalability remains constant during the evaluation period. However, in reality, resource scalability is dynamically affected by multiple factors, including the environment, user behavior, and device status.

[0013] Static evaluation methods cannot reflect the dynamic adjustment characteristics of resources, resulting in a large deviation between the evaluation results and the actual adjustable capacity. Summary of the Invention

[0014] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, one purpose of the present invention is to propose a method for quantitatively evaluating the adjustable capacity of light, storage, charging, and loading resources within a distribution network substation. This method achieves accurate quantification of the adjustable capacity of light, storage, charging, and loading resources through four steps: data collection, feature modeling, dynamic correction, and comprehensive evaluation. This method proposes a multi-resource collaborative modeling framework for the first time, combined with a real-time data-driven dynamic correction mechanism, which significantly improves the evaluation accuracy and practicality.

[0015] To this end, the present invention proposes a method for quantitatively evaluating the adjustable capacity of solar, storage, charging, and load resources within a distribution network area, comprising the following steps:

[0016] S1: Data collection: Collect PV output, energy storage SOC, charging pile charging demand and load power consumption data through smart terminals;

[0017] S2: Data preprocessing: The collected data were normalized using min-max normalization, Z-score standardization, and one-hot encoding methods;

[0018] S3: Feature modeling, including:

[0019] Photovoltaic resources use LSTM networks to predict future output ranges;

[0020] The energy storage resource calculates the adjustable capacity based on SOC and charge and discharge power limits;

[0021] Charging pile resources are constructed to assess the interruptible capacity using user response coefficients;

[0022] Load resources are categorized by type to quantify their adjustable potential;

[0023] S4: Dynamic scenario correction: Establish state space and action space through Markov decision process, and construct reward function by combining electricity price difference and equipment loss;

[0024] S5: Comprehensive evaluation: The entropy weight TOPSIS method is used to calculate the comprehensive score of the adjustable capacity of each resource and generate a substation regulation potential report.

[0025] Preferably, in S2:

[0026] The normalized formula for photovoltaic output is:

[0027]

[0028] Where: x is the original data value; x min is the minimum value of the data; x max is the maximum value of the data; x1 is the normalized value, ranging from [0, 1];

[0029] The formula for normalizing energy storage SOC is:

[0030]

[0031] Where: x is the original data value; μ is the mean of the data; σ is the standard deviation of the data; x2 is the standardized value, ranging from [3, 3];

[0032] The load type is represented by a three-digit one-hot encoding.

[0033] Preferably, in S3:

[0034] The calculation formula for photovoltaic adjustable capacity is:

[0035]

[0036] Where: is the output range predicted in the future based on the LSTM network, η PV is the adjustment efficiency coefficient, the default value is 0.95, ΔT is the adjustment time window, unit is h;

[0037] The calculation formula for the adjustable capacity of energy storage is:

[0038]

[0039] In the formula The capacity can be increased for energy storage. The capacity of energy storage can be adjusted downwards, E rated is the rated capacity in kWh, and Obtained through the technical manual or experimental data of the energy storage device, it reflects the maximum power limit of the energy storage device during the charging and discharging process.

[0040] Preferably, a Markov decision process design is required for dynamic scene correction in S4, including:

[0041] State space definition: {PV output, energy storage SOC, charging pile status, load status};

[0042] Action space definition: {adjust photovoltaic output, adjust energy storage charging and discharging, adjust charging pile power, and adjust load power consumption};

[0043] Construct a reward function: R = electricity price difference × regulated electricity volume - equipment loss cost - user compensation fee.

[0044] Preferably, when S5 performs a comprehensive assessment:

[0045] Entropy weight method is used to calculate the weight of each indicator: by calculating the information entropy of each resource indicator, its weight in the comprehensive evaluation is determined, where the information entropy of each indicator E j :

[0046]

[0047] Weight w j :

[0048]

[0049] TOPSIS method calculates proximity: by calculating the distance between each resource and the ideal solution and the negative ideal solution, reflecting its advantages and disadvantages in the comprehensive evaluation, constructing the standardized decision matrix Z, and determining the positive ideal solution Z + and negative ideal solution Z - ;

[0050] Calculate the closeness S of each resource to the ideal solution i :

[0051]

[0052] in, is the distance to the positive ideal solution, is the distance to the negative ideal solution.

[0053] The advantages of the present invention compared with the prior art are:

[0054] 1. Multi-resource collaborative modeling: For the first time, a unified and adjustable capacity quantification framework for four types of resources, namely solar, storage, charging, and load, is proposed, which overcomes the limitations of single resource assessment in existing methods.

[0055] 2. Dynamic scene adaptation: The introduction of a real-time data-driven dynamic correction mechanism significantly improves evaluation accuracy, outperforming traditional static evaluation methods.

[0056] 3. User behavior integration: Incorporating charging pile user elasticity and load classification characteristics into the evaluation system enhances the practicality of the method. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0058] Figure 1 This is a system architecture diagram of the present invention;

[0059] Figure 2 This is a flow chart of the dynamic prediction of photovoltaic adjustable capacity of the present invention;

[0060] Figure 3 This is a logic diagram for calculating the adjustable capacity of energy storage according to the present invention. DETAILED DESCRIPTION

[0061] The following describes in detail embodiments of the present application. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application and are not to be construed as limiting the present application.

[0062] In the description of this application, it should be noted that, unless otherwise expressly specified or limited, the terms "installed," "connected," and "connected" should be understood in a broad sense. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; they can refer to internal communication between two components or the interaction between two components. Those skilled in the art will understand the specific meanings of the above terms in this application based on the specific circumstances.

[0063] The present invention will be described in further detail below with reference to the accompanying drawings.

[0064] Combine Figures 1 to 3 The present invention proposes a method for quantitatively evaluating the adjustable capacity of light, storage, charging and load resources in a distribution network area, comprising the following steps:

[0065] S1: Data collection: Collect PV output, energy storage SOC, charging pile charging demand and load power consumption data through smart terminals;

[0066] S2: Data preprocessing: The collected data were normalized using min-max normalization, Z-score standardization, and one-hot encoding methods;

[0067] S3: Feature modeling, including:

[0068] Photovoltaic resources use LSTM networks to predict future output ranges;

[0069] The energy storage resource calculates the adjustable capacity based on SOC and charge and discharge power limits;

[0070] Charging pile resources are constructed to assess the interruptible capacity using user response coefficients;

[0071] Load resources are categorized by type to quantify their adjustable potential;

[0072] S4: Dynamic scenario correction: Establish state space and action space through Markov decision process, and construct reward function by combining electricity price difference and equipment loss;

[0073] S5: Comprehensive evaluation: The entropy weight TOPSIS method is used to calculate the comprehensive score of the adjustable capacity of each resource and generate a substation regulation potential report.

[0074] Preferably, in S2:

[0075] The normalized formula for photovoltaic output is:

[0076]

[0077] Where: x is the original data value; x min is the minimum value of the data; x max is the maximum value of the data; x1 is the normalized value, ranging from [0, 1];

[0078] The formula for normalizing energy storage SOC is:

[0079]

[0080] Where: x is the original data value; μ is the mean of the data; σ is the standard deviation of the data; x2 is the standardized value, ranging from [3, 3];

[0081] The load type is represented by a three-digit one-hot encoding.

[0082] Preferably, in S3:

[0083] The calculation formula for photovoltaic adjustable capacity is:

[0084]

[0085] Where: is the output range predicted in the future based on the LSTM network, η PV is the adjustment efficiency coefficient, the default value is 0.95, ΔT is the adjustment time window, unit is h;

[0086] The calculation formula for the adjustable capacity of energy storage is:

[0087]

[0088] In the formula The capacity can be increased for energy storage. The capacity of energy storage can be adjusted downwards, E rated is the rated capacity in kWh, and Obtained through the technical manual or experimental data of the energy storage device, it reflects the maximum power limit of the energy storage device during the charging and discharging process.

[0089] Preferably, a Markov decision process design is required for dynamic scene correction in S4, including:

[0090] State space definition: {PV output, energy storage SOC, charging pile status, load status};

[0091] Action space definition: {adjust photovoltaic output, adjust energy storage charging and discharging, adjust charging pile power, and adjust load power consumption};

[0092] Construct a reward function: R = electricity price difference × regulated electricity volume - equipment loss cost - user compensation fee.

[0093] Preferably, when S5 performs a comprehensive assessment:

[0094] Entropy weight method is used to calculate the weight of each indicator: by calculating the information entropy of each resource indicator, its weight in the comprehensive evaluation is determined, where the information entropy of each indicator E j :

[0095]

[0096] Weight w j :

[0097]

[0098] TOPSIS method calculates proximity: by calculating the distance between each resource and the ideal solution and the negative ideal solution, reflecting its advantages and disadvantages in the comprehensive evaluation, constructing the standardized decision matrix Z, and determining the positive ideal solution Z + and negative ideal solution Z - ;

[0099] Calculate the closeness S of each resource to the ideal solution i :

[0100]

[0101] in, is the distance to the positive ideal solution, is the distance to the negative ideal solution.

[0102] In order to more clearly illustrate the specific embodiment of the present invention, an embodiment is provided below:

[0103] The present invention proposes a method for quantitatively evaluating the adjustable capacity of solar, storage, charging, and load resources within a distribution network area, including:

[0104] 1. Data collection and preprocessing:

[0105] Real-time data collection is performed through intelligent terminals (TTUs), smart meters, meteorological sensors and other equipment:

[0106] Photovoltaic: light intensity, temperature, photovoltaic output, and photovoltaic historical output curve.

[0107] Energy storage: SOC (state of charge), charge and discharge efficiency, rated capacity, and charge and discharge power limit.

[0108] Charging pile: charging demand period, user reservation information, and maximum interruption time.

[0109] Load: time-of-use electricity consumption, load type (interruptible / transferable / rigid).

[0110] 2. Data normalization processing:

[0111] 1. Normalization method selection

[0112] According to different types of data characteristics, the following normalization methods are used:

[0113] (1) Minimum and maximum normalization: Applicable to data with clear upper and lower limits (such as photovoltaic output, SOC, charging power, etc.).

[0114]

[0115] Where:

[0116] x is the original data value; x min is the minimum value of the data; x max is the maximum value of the data; x1 is the normalized value, usually in the range of [0, 1].

[0117] (2) Zscore normalization: Applicable to data that is greatly affected by environmental factors (such as temperature).

[0118]

[0119] in:

[0120] x is the original data value; μ is the mean of the data; σ is the standard deviation of the data; x2 is the standardized value, with no fixed range, usually [3, 3].

[0121] (3) One-hot encoding: used for categorical variables (such as load type). For categorical variables such as load type, one-hot encoding is used. For example, the three load types (interruptible, transferable, and rigid) can be encoded as a three-bit binary vector, such as [1, 0, 0] for interruptible load, [0, 1, 0] for transferable load, and [0, 0, 1] for rigid load.

[0122] 2. Parameter normalization rules

[0123]

[0124]

[0125] 3. Adjustable Capability Feature Modeling:

[0126] 1. Photovoltaic resource adjustable capacity model

[0127] Input features: historical output curve, light intensity, temperature, and time series (hourly level).

[0128] Prediction method: Based on LSTM network (long short-term memory network) to predict the output range in the future period The LSTM network consists of multiple hidden layers, each containing several LSTM units. Its input is historical PV output data, and its output is a range of PV output for future periods. During training, the model is trained using historical data, and model parameters are adjusted through cross-validation.

[0129] Adjustable capacity formula:

[0130]

[0131] Among them, η PV is the adjustment efficiency coefficient (default is 0.95), and ΔT is the adjustment time window (hours).

[0132] (2) Energy storage resource adjustable capacity model

[0133] Calculate the energy storage capacity that can be increased based on the current SOC and charge and discharge power limits

[0134] With adjustable capacity

[0135]

[0136] Among them, E rated is the rated capacity (kWh), and Obtained through the technical manual or experimental data of the energy storage device, it reflects the maximum power limit of the energy storage device during the charging and discharging process.

[0137] (3) Charging pile adjustable capacity model

[0138] Based on the elasticity of user charging demand, an interruptible capacity evaluation function is constructed:

[0139]

[0140] Among them, α i is the user response coefficient, t delay,i The maximum delay time.

[0141] (4) Load Adjustability Model

[0142] Quantification by load type:

[0143] Interruptible load: Directly calculate the maximum curtailable power and analyze historical electricity consumption data to determine the maximum curtailment amount of interruptible load within a specific period.

[0144] Transferable load: The time-shift potential is fitted based on historical data. By analyzing historical electricity consumption data, the changes in electricity consumption of transferable loads in different time periods are determined, thereby evaluating their time-shift potential.

[0145] 4. Dynamic scene correction mechanism

[0146] 1. Markov Decision Process (MDP) Design

[0147] State space: includes current photovoltaic output, energy storage SOC, charging pile status, load status, etc.

[0148] Action space: including adjusting photovoltaic output, adjusting energy storage charging and discharging power, adjusting charging pile charging power, adjusting load power consumption, etc.

[0149] Reward function:

[0150] R = Adjustment income (electricity price difference) Adjustment cost (equipment loss, user compensation)

[0151] 2. Multi-timescale assessment framework

[0152] Minute-level evaluation: used for real-time adjustment, reflecting the system's adjustment capabilities in a short period of time (such as energy storage charging and discharging).

[0153] Hourly assessment: used for short-term scheduling, reflecting the system's regulatory potential over a longer period of time (such as load transfer).

[0154] Day-ahead assessment: used for medium- and long-term planning, reflecting the system's regulation capabilities within the next day (such as photovoltaic output forecast).

[0155] 5. Comprehensive Adjustability Assessment

[0156] 1. Entropy weight TOPSIS method:

[0157] Entropy weight method: By calculating the information entropy of each resource indicator, its weight in the comprehensive evaluation is determined.

[0158] Calculate the information entropy E of each indicator j :

[0159]

[0160] Weight w j :

[0161]

[0162] TOPSIS method: By calculating the distance between each resource and the ideal solution and the negative ideal solution, it reflects its advantages and disadvantages in the comprehensive evaluation. Construct a standardized decision matrix Z and determine the positive ideal solution Z + and negative ideal solution Z - .

[0163] Calculate the closeness S of each resource to the ideal solution i :

[0164]

[0165] in, is the distance to the positive ideal solution, is the distance to the negative ideal solution.

[0166] 2. Substation resource adjustment capacity indicator matrix:

[0167]

[0168] The entropy weight TOPSIS method is used to calculate the comprehensive score of the adjustable capacity of each resource, generate an overall regulation potential report for the substation area, and clarify the regulation priority of each resource (such as energy storage > photovoltaic > load > charging pile).

[0169] Example:

[0170] 1. Select a substation (including 500kW photovoltaic power generation, 200kWh energy storage, 20 charging piles, and 300kW adjustable load).

[0171] 2. Real-time meteorological data is collected and the photovoltaic output range is predicted to be [80kW, 350kW].

[0172] 3. Calculate based on energy storage SOC = 60%

[0173] 4. The evaluation results show that the overall upward adjustment capacity of the substation is 720kWh and the downward adjustment capacity is 480kWh.

[0174] The present invention and its embodiments are described above. This description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if a person skilled in the art is inspired by this and, without departing from the purpose of the present invention, designs structures and embodiments similar to this technical solution without inventiveness, they shall fall within the scope of protection of the present invention.

Claims

1. A method for quantitatively evaluating the adjustable capacity of solar, storage, charging, and load resources within a distribution network area, characterized in that: The following steps are involved: S1: Data collection: Collecting PV output, energy storage SOC, charging pile charging demand and load power consumption data through smart terminals; S2: Data preprocessing: The collected data were normalized using min-max normalization, Z-score standardization, and one-hot encoding methods; S3: Feature modeling, including: Photovoltaic resources use LSTM networks to predict future output ranges; The energy storage resource calculates the adjustable capacity based on SOC and charge and discharge power limits; Charging pile resources are constructed to assess the interruptible capacity using user response coefficients; Load resources are classified by type to quantify their adjustable potential; S4: Dynamic scenario correction: Establish state space and action space through Markov decision process, and construct reward function by combining electricity price difference and equipment loss; S5: Comprehensive evaluation: Use the entropy weight TOPSIS method to calculate the comprehensive score of the adjustable capacity of each resource and generate a substation regulation potential report.

2. A method for quantitatively evaluating the adjustable capacity of solar, storage, charging, and load resources within a distribution network area according to claim 1, characterized in that: In S2: The normalized formula for photovoltaic output is: Where: x is the original data value; x min is the minimum value of the data; x max is the maximum value of the data; x1 is the normalized value, ranging from [0, 1]; The formula for normalizing energy storage SOC is: Where: x is the original data value; μ is the mean of the data; σ is the standard deviation of the data; x2 is the standardized value, ranging from [3, 3]; The load type is represented by a three-digit one-hot encoding.

3. The method for quantitatively evaluating the adjustable capacity of solar, storage, charging, and load resources within a distribution network area according to claim 1 is characterized by: In the S3: The calculation formula for photovoltaic adjustable capacity is: Where: is the output range predicted in the future based on the LSTM network, η PV is the adjustment efficiency coefficient, the default value is 0.95, ΔT is the adjustment time window, unit is h; The calculation formula for the adjustable capacity of energy storage is: In the formula The capacity can be increased for energy storage. The capacity of energy storage can be adjusted downwards, E rated is the rated capacity kWh, and Obtained through the technical manual or experimental data of the energy storage device, it reflects the maximum power limit of the energy storage device during the charging and discharging process.

4. The method for quantitatively evaluating the adjustable capacity of solar, storage, charging, and load resources within a distribution network area according to claim 1 is characterized by: The dynamic scene correction in S4 requires a Markov decision process design, including: State space definition: {PV output, energy storage SOC, charging pile status, load status}; Action space definition: {adjust photovoltaic output, adjust energy storage charging and discharging, adjust charging pile power, and adjust load power consumption}; Construct a reward function: R = electricity price difference × regulated electricity volume - equipment loss cost - user compensation fee.

5. The method for quantitatively evaluating the adjustable capacity of solar, storage, charging, and load resources within a distribution network area according to claim 1 is characterized by: The S5 comprehensive assessment is: Entropy weight method is used to calculate the weight of each indicator: by calculating the information entropy of each resource indicator, its weight in the comprehensive evaluation is determined, where the information entropy of each indicator E j : Weight w j : TOPSIS method calculates proximity: by calculating the distance between each resource and the ideal solution and the negative ideal solution, reflecting its advantages and disadvantages in the comprehensive evaluation, constructing the standardized decision matrix Z, and determining the positive ideal solution Z + and negative ideal solution Z - ; Calculate the closeness S of each resource to the ideal solution i : in, is the distance to the positive ideal solution, is the distance to the negative ideal solution.

Images