Power transaction settlement model construction method based on virtual power plant

Abstract

The invention discloses an electric power transaction settlement model construction method based on a virtual power plant, and relates to the technical field of electric power transaction settlement, a dynamic settlement period adjustment and sliding filtering linkage mechanism is adopted, and electric quantity oscillation distortion caused by high-frequency charging and discharging is effectively suppressed while market price response sensitivity is reserved; equipment physical loss and market credit risks are quantified into executable decision parameters, and the problem that a traditional model excessively pursues economical efficiency and neglects the service life of equipment is solved; an off-line compensation strategy of topological relation and power consumption mode similarity is fused, data integrity is maintained in a weak network environment, and settlement disputes caused by communication interruption are avoided; in addition, a health degree compensation mechanism of block chain verification ensures transparent allocation of equipment aging cost among transaction parties, and promotes market fairness.

Description

Technical Field

[0001] The present invention relates to the technical field of power transaction settlement, and in particular to a method for constructing a power transaction settlement model based on a virtual power plant. Background Art

[0002] The current mainstream model for virtual power plants involves aggregating distributed energy storage systems to participate in multi-tiered electricity market transactions. Driven by the high-frequency fluctuations of the spot market and the sub-second response of frequency regulation ancillary services, energy storage devices must implement minute-by-minute charging and discharging strategy switching to capture arbitrage opportunities. This dynamic strategy leads to a surge in the frequency of charging and discharging operations, causing the net power recorded by the metering system to fluctuate violently within the settlement cycle. During peak and off-peak electricity price switching periods, energy storage units often complete multiple charge and discharge state transitions within 15 minutes, blurring the trading platform's understanding of the actual delivered power.

[0003] Newer solutions primarily use reinforcement learning to optimize charging and discharging strategies and introduce blockchain technology for dynamic settlement. For example, the DRL-EnergyScheduler algorithm deployed by some provincial trading platforms dynamically adjusts charging and discharging thresholds through Q-learning. Other projects use the Hyperledger Fabric architecture to record energy storage status changes with a 5-minute granularity. However, in these cases, the reinforcement learning model overly pursues economic optimization and ignores the capacitor loss costs caused by frequent device switching. Furthermore, the blockchain timestamp mechanism and physical device state transitions have millisecond-level deviations, resulting in duplicate measurement of the same electricity consumption across different settlement dimensions.

[0004] Some solutions introduce sliding window filtering algorithms to smooth the charge and discharge curves and establish equipment health compensation factors; for example, the Switching-Cost evaluation model quantifies the loss of capacitor cycle life as a settlement correction item; another team proposes a dynamic settlement cycle adjustment mechanism that automatically extends the metering window to 30 minutes when high-frequency switching is detected; however, these methods give rise to new problems: the filtering algorithm weakens the real-time transmission effect of market prices, and the dynamic cycle adjustment causes disputes over the fairness of user revenue distribution. Summary of the Invention

[0005] In view of the above existing problems, the present invention is proposed.

[0006] The present invention provides a method for constructing an electricity trading settlement model based on a virtual power plant to solve the problem that existing solutions rely on reinforcement learning optimization strategies and blockchain settlement, but have defects such as high-frequency switching error accumulation, unquantified capacitor loss, and deviation between timestamp and physical state, making it difficult to balance economy and equipment life.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] The embodiment of the present invention provides a method for constructing an electricity transaction settlement model based on a virtual power plant, which includes:

[0009] Step S1: collecting the charging and discharging status data of distributed energy storage devices, real-time price data of the electricity market, and equipment operating parameters in real time through edge computing nodes;

[0010] Step S2: constructing a charge-discharge strategy decision model to calculate a switching cost threshold including capacitor loss cost and market default risk before each charge-discharge state switch;

[0011] Step S3: Dynamically adjust the settlement period length based on the current electricity price fluctuation rate and equipment health. When it is detected that the number of charge and discharge actions within a preset period exceeds a critical value, extend the current settlement period to 1.2-1.8 times the original period.

[0012] The triggering conditions for dynamically adjusting the settlement period in step S3 include:

[0013] It is detected that the number of charge and discharge actions exceeds 150% of the preset baseline value within three consecutive settlement cycles;

[0014] The electricity market price volatility during the current period exceeds 2 standard deviations of the previous day's price volatility during the same period;

[0015] The device health indicator drops below 80% of the initial value;

[0016] Methods for generating preset baseline values include:

[0017] Take the moving average of the number of charge and discharge operations in the same period of the previous 7 days;

[0018] The recommended value of rated operating frequency provided by the superimposed equipment manufacturer;

[0019] Introducing holiday factors for seasonal correction;

[0020] Step S4, performing sliding window filtering on the net power data stream generated by high-frequency charging and discharging to generate a smoothed settlement power sequence;

[0021] Step S5: Establish a device health decay model, quantify the capacitor cycle life loss as a settlement correction coefficient, and embed it into the transaction settlement logic;

[0022] Step S6: In the case of communication interruption, offline data compensation is performed based on the similarity of power consumption patterns and topological relationships of adjacent nodes;

[0023] Step S7: Perform multi-level matching on the processed settlement data with the power trading platform to generate a traceable transaction certificate chain.

[0024] As a preferred solution of the method for constructing an electricity trading settlement model based on a virtual power plant described in the present invention, the edge computing node in step S1 includes a data acquisition module, a protocol conversion module and a local cache module, wherein the protocol conversion module supports parallel parsing of Modbus, IEC61850 and MQTT protocols.

[0025] As a preferred solution of the method for constructing a power transaction settlement model based on a virtual power plant according to the present invention, the calculation of the switching cost threshold in step S2 includes:

[0026] Calculate the capacitance loss cost of a single action based on the rated number of cycles and current health of the energy storage device;

[0027] Assess market default risk weights based on historical dispatch order execution rates;

[0028] The cost and the fluctuation range of electricity prices in the current period are weighted and integrated to generate a dynamic threshold.

[0029] As a preferred solution of the method for constructing a power transaction settlement model based on a virtual power plant according to the present invention, in step S2, the steps of constructing a charge and discharge strategy decision model are as follows:

[0030] At time t, the charge and discharge action is represented as a binary control variable u t ∈{+1,-1}, where +1 represents charging and -1 represents discharging. The model determines the optimal strategy by maximizing the cumulative net benefit over the entire scheduling period. The formula is:

[0031] in,

[0032]

[0033] Among them, u t represents the charge and discharge control variable at time t, p t represents the electricity price in period t, ΔE t Indicates the charge or discharge energy in time period t, ΔC t Represents the equipment loss cost caused by charging and discharging, R t represents the market default risk cost, C b Represents the unit replacement cost of energy storage equipment, N r Indicates the rated number of cycles of the equipment, SoH t represents the health of the device at time t, ω t represents the market default risk weight, η t Indicates the scheduling instruction execution rate, n exec Indicates the number of historical execution instructions, n order Indicates the total number of historical instructions issued, |ΔEt | represents the absolute value of the charge and discharge energy, and T represents the number of time steps in the entire scheduling period;

[0034] In step S2, the step of calculating the switching cost threshold is:

[0035] Before each charge and discharge state switch, a dynamic threshold is used to measure the switching cost:

[0036] Among them, Θ t represents the switching cost threshold at time t, α, β, and γ represent the weighted coefficients of loss cost, market default risk, and electricity price fluctuation range, respectively, and α+β+γ=1. represents the average electricity price within the rolling window, Indicates the deviation of the electricity price from the average electricity price.

[0037] As a preferred solution of the method for constructing a power transaction settlement model based on a virtual power plant described in the present invention, in step S3, during the process of dynamically adjusting the length of the settlement period, the length of the settlement period is dynamically determined through a linkage mechanism between electricity price volatility and equipment health, and the current period is extended when the number of actions exceeds a threshold;

[0038] The electricity price volatility is defined as:

[0039]

[0040] Among them, σ t represents the electricity price volatility at time t, W represents the time window length used to calculate the volatility, and P i represents the electricity price in period i, represents the average electricity price within the window;

[0041] The normalized volatility is defined as:

[0042]

[0043] Among them, t represents the normalized electricity price volatility, σ ref represents the reference volatility threshold;

[0044] The formula for calculating the length of the basic settlement cycle is:

[0045] L t =L0(1+α3ζ t +α4(1-SoH t )),

[0046] Among them, L trepresents the length of the basic settlement period at time t, L0 represents the length of the benchmark period, α3 represents the weight of the impact of electricity price fluctuations, α4 represents the weight of the impact of health, and SoH t Indicates the health of the device at time t;

[0047] Introduce an over-limit extension mechanism, specifically:

[0048]

[0049] Among them, L' t Indicates the length of the extended settlement cycle, n act Indicates the total number of charge and discharge actions within the monitoring period U, N crit It represents the critical value of the number of actions, β represents the over-limit extension coefficient, and U represents the length of the monitoring period.

[0050] As a preferred solution of the method for constructing a power transaction settlement model based on a virtual power plant described in the present invention, the sliding window filtering process in step S4 adopts a variable window width mechanism, and the window width is dynamically adjusted according to the following parameters:

[0051] The current price change gradient of the electricity market;

[0052] The percentage of remaining available capacity of the energy storage device;

[0053] The electricity consumption deviation rate between adjacent settlement cycles.

[0054] As a preferred solution of the method for constructing a power transaction settlement model based on a virtual power plant according to the present invention, in step S4, the sliding window filtering process is performed as follows:

[0055] At time t, based on the price difference between the current period and the previous period, calculate the electricity price change gradient:

[0056] G t =|P t -P t-1 |,

[0057] Among them, G t represents the gradient of electricity price change at time t, P t represents the electricity price in period t, P t-1 represents the electricity price in period t-1;

[0058] Calculate the electricity deviation rate of adjacent cycles to measure the relative difference in net electricity between two adjacent settlement cycles. The deviation rate is defined as:

[0059]

[0060] Among them, D t Indicates the power deviation rate between the current settlement period and the previous period. Indicates the net electricity consumption of the last settlement period. Indicates the net electricity consumption of the previous two cycles, and n indicates the sequence number of the current settlement cycle;

[0061] Calculate the dynamic window width, set the base window width as W0, and combine the electricity price gradient, remaining capacity and cycle deviation rate to construct the dynamic window width:

[0062]

[0063] Among them, W t represents the sliding window width at time t, W0 represents the reference window width, α5 represents the influence weight of the electricity price gradient, α6 represents the influence weight of the remaining capacity, and SoC t represents the percentage of remaining available capacity at time t, α7 represents the impact weight of the cycle deviation rate, Indicates rounding down, clip(x,a,b) means limiting x to the interval [a,b], W min With W max Represents the minimum and maximum window width respectively;

[0064] The high-frequency net power data x is processed by a dynamic width sliding window t Perform average filtering and output the smoothed settlement power sequence:

[0065]

[0066] Among them, y t represents the smoothed settlement electricity at time t, x i Represents the net power data at the i-th moment.

[0067] As a preferred solution of the method for constructing a power transaction settlement model based on a virtual power plant according to the present invention, the equipment health decay model in step S5 includes:

[0068] Capacitance decay curve fitting module, which establishes a three-dimensional mapping relationship based on the device operating temperature, charge and discharge depth, and cycle number;

[0069] The health compensation coefficient generation module converts the capacity decay rate into a settlement amount correction ratio;

[0070] The compensation coefficient verification module achieves multi-party consensus confirmation through blockchain smart contracts.

[0071] As a preferred solution of the method for constructing a power transaction settlement model based on a virtual power plant according to the present invention, the specific implementation of the offline data compensation in step S6 includes:

[0072] Build a user electricity consumption pattern feature vector library, including typical electricity consumption curves and abnormal pattern labels;

[0073] The power consumption similarity weights between disconnected nodes and adjacent nodes are calculated based on the graph attention network;

[0074] Generate compensation power estimates based on topological distance and similarity weights.

[0075] As a preferred solution of the method for constructing a power transaction settlement model based on a virtual power plant according to the present invention, the power consumption pattern feature vector library includes:

[0076] Fourier transform low-frequency component coefficients of a typical daily load curve;

[0077] Sliding window statistics of the proportion of peak and valley periods in electricity consumption;

[0078] The marking frequency of abnormal power usage events, such as power sag and sustained overload.

[0079] The beneficial effects of the present invention are as follows: the dynamic settlement cycle adjustment and sliding filter linkage mechanism of the present invention effectively suppresses the power oscillation distortion caused by high-frequency charging and discharging while retaining the market price response sensitivity; quantifies the physical loss of equipment and market credit risk into executable decision parameters, breaking the dilemma of traditional models that excessively pursue economy while ignoring equipment life; integrates the offline compensation strategy of topological relationship and power consumption pattern similarity to maintain data integrity in a weak network environment and avoid settlement disputes caused by communication interruption; in addition, the health compensation mechanism verified by blockchain ensures that the cost of equipment aging is transparently shared among the parties to the transaction, thereby promoting market fairness.

[0080] The present invention takes into account both real-time response and long-term sustainability, forming a digital twin mapping system of physical device status and virtual transaction rules. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. 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 any creative work.

[0082] Figure 1 This is a flow chart of the method for constructing an electricity transaction settlement model based on a virtual power plant in Example 1. DETAILED DESCRIPTION

[0083] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0084] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0085] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0086] Example 1, with reference to Figure 1 This embodiment provides a method for constructing a power transaction settlement model based on a virtual power plant, comprising the following steps:

[0087] Step S1: collecting the charging and discharging status data of distributed energy storage devices, real-time price data of the electricity market, and equipment operating parameters in real time through edge computing nodes;

[0088] In step S1, the edge computing node includes a data acquisition module, a protocol conversion module, and a local cache module, wherein the protocol conversion module supports parallel parsing of Modbus, IEC61850, and MQTT protocols;

[0089] Step S2: constructing a charge-discharge strategy decision model to calculate a switching cost threshold including capacitor loss cost and market default risk before each charge-discharge state switch;

[0090] The calculation of the handover cost threshold in step S2 includes:

[0091] Calculate the capacitance loss cost of a single action based on the rated number of cycles and current health of the energy storage device;

[0092] Assess market default risk weights based on historical dispatch order execution rates;

[0093] The cost and the fluctuation range of electricity price in the current period are weighted and integrated to generate dynamic thresholds;

[0094] In step S2, the steps of constructing the charge and discharge strategy decision model are as follows:

[0095] At time t, the charge and discharge action is represented as a binary control variable u t ∈{+1,-1}, where +1 represents charging and -1 represents discharging. The model determines the optimal strategy by maximizing the cumulative net benefit over the entire scheduling period. The formula is:

[0096] in,

[0097]

[0098] Among them, u t represents the charge and discharge control variable at time t, P t represents the electricity price in period t, ΔE t Indicates the charge or discharge energy in time period t, ΔC t Represents the equipment loss cost caused by charging and discharging, R t represents the market default risk cost, C b represents the unit replacement cost of energy storage equipment, N r Indicates the rated number of cycles of the equipment, SoH t represents the health of the device at time t, ω t represents the market default risk weight, η t Indicates the scheduling instruction execution rate, n exec Indicates the number of historical execution instructions, n order Indicates the total number of historical instructions issued, |ΔE t | represents the absolute value of the charge and discharge energy, and T represents the number of time steps in the entire scheduling period;

[0099] In step S2, the steps for calculating the switching cost threshold are:

[0100] Before each charge and discharge state switch, a dynamic threshold is used to measure the switching cost:

[0101] Among them, Θ t represents the switching cost threshold at time t, α, β, and γ represent the weighted coefficients of loss cost, market default risk, and electricity price fluctuation range, respectively, and α+β+γ=1. represents the average electricity price within the rolling window, Indicates the deviation of electricity price from the average electricity price;

[0102] Specifically, the decision-making model integrates the economic benefits of charging and discharging actions with equipment loss and market credit risk. It uses binary control variables to clarify action selection. The loss cost model combines replacement cost, cycle life, and health to truly reflect the relationship between equipment aging and usage. The risk cost is calculated based on historical execution rates to provide credit risk guidance for the strategy. The dynamic threshold integrates these three factors in a weighted manner, flexibly adjusting the action switching threshold under different market fluctuations to effectively balance benefits and risks.

[0103] Step S3: Dynamically adjust the settlement period length based on the current electricity price fluctuation rate and equipment health. When it is detected that the number of charge and discharge actions within a preset period exceeds a critical value, extend the current settlement period to 1.2-1.8 times the original period.

[0104] The triggering conditions for dynamically adjusting the settlement period in step S3 include:

[0105] It is detected that the number of charge and discharge actions exceeds 150% of the preset baseline value within three consecutive settlement cycles;

[0106] The electricity market price volatility during the current period exceeds 2 standard deviations of the previous day's price volatility during the same period;

[0107] The device health indicator drops below 80% of the initial value;

[0108] Methods for generating preset baseline values include:

[0109] Take the moving average of the number of charge and discharge operations in the same period of the previous 7 days;

[0110] The recommended value of rated operating frequency provided by the superimposed equipment manufacturer;

[0111] Introducing holiday factors for seasonal correction;

[0112] In step S3, during the process of dynamically adjusting the length of the settlement period, the length of the settlement period is dynamically determined through the linkage mechanism between electricity price fluctuation rate and equipment health, and the current period is extended when the number of actions exceeds the threshold;

[0113] The electricity price volatility is defined as:

[0114]

[0115] Among them, σ t represents the electricity price volatility at time t, W represents the time window length used to calculate the volatility, and P i represents the electricity price in period i, represents the average electricity price within the window;

[0116] The normalized volatility is defined as:

[0117]

[0118] Among them, t represents the normalized electricity price volatility, σ ref represents the reference volatility threshold;

[0119] The formula for calculating the length of the basic settlement cycle is:

[0120] L t =L0(1+α3ζ t +α4(1-SoH t )),

[0121] Among them, L trepresents the length of the basic settlement period at time t, L0 represents the length of the benchmark period, α3 represents the influence weight of electricity price fluctuation, α4 represents the influence weight of health, and SoH t Indicates the health of the device at time t;

[0122] Introducing an over-limit extension mechanism, specifically:

[0123]

[0124] Among them, L' t Indicates the length of the extended settlement cycle, n act Indicates the total number of charge and discharge actions within the monitoring period U, N crit It represents the critical value of the number of actions, β represents the over-limit extension coefficient, and U represents the length of the monitoring period;

[0125] Specifically, by synchronizing the settlement cycle with market fluctuations and equipment health, real-time adaptation of the cycle length is achieved. When electricity price volatility increases or equipment health decreases, the basic cycle can be shortened or extended accordingly, taking into account both accounting precision and system stability. The strategy of extending the number of actions beyond the limit can effectively prevent data noise and equipment fatigue caused by excessive settlement, thereby reducing operation and maintenance costs.

[0126] Step S4, performing sliding window filtering on the net power data stream generated by high-frequency charging and discharging to generate a smoothed settlement power sequence;

[0127] The sliding window filtering process in step S4 adopts a variable window width mechanism, and the window width is dynamically adjusted according to the following parameters:

[0128] The current price gradient of the electricity market;

[0129] The percentage of remaining available capacity of the energy storage device;

[0130] The electricity consumption deviation rate in adjacent settlement cycles;

[0131] In step S4, the sliding window filtering process is performed as follows:

[0132] At time t, based on the price difference between the current period and the previous period, calculate the price change gradient:

[0133] G t =|P t -P t-1 |,

[0134] Among them, G t represents the electricity price change gradient at time t, P t represents the electricity price in period t, P t-1 represents the electricity price in period t-1;

[0135] Calculate the electricity deviation rate of adjacent cycles to measure the relative difference in net electricity between two adjacent settlement cycles. The deviation rate is defined as:

[0136]

[0137] Among them, D t Indicates the power deviation rate between the current settlement period and the previous period. Indicates the net electricity consumption of the last settlement period. Indicates the net electricity consumption of the previous two cycles, and n indicates the sequence number of the current settlement cycle;

[0138] Calculate the dynamic window width, set the base window width as W0, and combine the electricity price gradient, remaining capacity and cycle deviation rate to construct the dynamic window width:

[0139]

[0140] Among them, W t represents the sliding window width at time t, W0 represents the reference window width, α5 represents the influence weight of the electricity price gradient, α6 represents the influence weight of the remaining capacity, and SoC t represents the percentage of remaining available capacity at time t, α7 represents the impact weight of the cycle deviation rate, Indicates rounding down, clip(x,a,b) means limiting x to the interval [a,b], W min With W max Represents the minimum and maximum window width respectively;

[0141] The high-frequency net power data x is processed by a dynamic width sliding window t Perform average filtering and output the smoothed settlement power sequence:

[0142]

[0143] Among them, y t represents the smoothed settlement electricity at time t, x i Represents the net power data at the i-th moment;

[0144] Specifically, the dynamic sliding window filter adaptively adjusts the filter width using three-dimensional indicators: real-time electricity price gradient, remaining capacity, and cycle deviation rate. This allows the high-frequency net power sequence to smooth out noise while retaining trend characteristics. When electricity prices fluctuate drastically or capacity decreases, the window is automatically widened to enhance the filtering effect. When cycle deviations are significant, the deviation rate is included in the width calculation to ensure data continuity and consistency. Parameters are updated online without manual intervention, balancing algorithm flexibility and settlement accuracy.

[0145] Step S5: Establish a device health decay model, quantify the capacitor cycle life loss as a settlement correction coefficient, and embed it into the transaction settlement logic;

[0146] The device health decay model in step S5 includes:

[0147] Capacitance decay curve fitting module, which establishes a three-dimensional mapping relationship based on the device operating temperature, charge and discharge depth, and cycle number;

[0148] The health compensation coefficient generation module converts the capacity decay rate into a settlement amount correction ratio;

[0149] The compensation coefficient verification module achieves multi-party consensus confirmation through blockchain smart contracts;

[0150] Step S6: In the case of communication interruption, offline data compensation is performed based on the similarity of power consumption patterns and topological relationships of adjacent nodes;

[0151] The specific implementation of offline data compensation in step S6 includes:

[0152] Build a user electricity consumption pattern feature vector library, including typical electricity consumption curves and abnormal pattern labels;

[0153] The power consumption similarity weights between disconnected nodes and adjacent nodes are calculated based on the graph attention network;

[0154] Generate compensation power estimation value based on topological distance and similarity weight;

[0155] The power consumption pattern feature vector library includes:

[0156] Fourier transform low-frequency component coefficients of a typical daily load curve;

[0157] Sliding window statistics of the proportion of peak and valley periods in electricity consumption;

[0158] The frequency of abnormal power usage events, including power sags and sustained overloads;

[0159] Step S7: Perform multi-level matching on the processed settlement data with the power trading platform to generate a traceable transaction certificate chain.

[0160] 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 the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for constructing an electricity transaction settlement model based on a virtual power plant, characterized in that: include, Step S1: collecting the charging and discharging status data of distributed energy storage devices, real-time price data of the electricity market, and equipment operating parameters in real time through edge computing nodes; Step S2: constructing a charge-discharge strategy decision model to calculate a switching cost threshold including capacitor loss cost and market default risk before each charge-discharge state switch; Step S3: Dynamically adjust the length of the settlement period based on the current electricity price fluctuation rate and the health of the equipment. When it is detected that the number of charge and discharge actions within the preset period exceeds a critical value, the current settlement period is extended. Step S4, performing sliding window filtering on the net power data stream generated by high-frequency charging and discharging to generate a smoothed settlement power sequence; Step S5: Establish a device health decay model, quantify the capacitor cycle life loss as a settlement correction coefficient, and embed it into the transaction settlement logic; Step S6: In the case of communication interruption, offline data compensation is performed based on the similarity of power consumption patterns and topological relationships of adjacent nodes; Step S7: Perform multi-level matching on the processed settlement data with the power trading platform to generate a traceable transaction certificate chain.

2. A method for constructing a power transaction settlement model based on a virtual power plant according to claim 1, characterized in that: The edge computing node in step S1 includes a data acquisition module, a protocol conversion module and a local cache module, wherein the protocol conversion module supports parallel parsing of Modbus, IEC61850 and MQTT protocols.

3. The method for constructing a power transaction settlement model based on a virtual power plant according to claim 1, characterized in that: The calculation of the handover cost threshold in step S2 includes: Calculate the capacitance loss cost of a single action based on the rated number of cycles and current health of the energy storage device; Assess market default risk weights based on historical dispatch order execution rates; The cost and the fluctuation range of electricity prices in the current period are weighted and integrated to generate a dynamic threshold.

4. A method for constructing a power transaction settlement model based on a virtual power plant according to claim 3, characterized in that: In step S2, the steps of constructing the charge and discharge strategy decision model are: At time t, the charge and discharge action is represented as a binary control variable u t ∈{+1,-1}, where +1 represents charging and -1 represents discharging. The model determines the optimal strategy by maximizing the cumulative net benefit over the entire scheduling period. The formula is: in, R t =ω t |NO t |,oh t =1st t , Among them, u t represents the charge and discharge control variable at time t, P t represents the electricity price in period t, ΔE t Indicates the charge or discharge energy in time period t, ΔC t Represents the equipment loss cost caused by charging and discharging, R t represents the market default risk cost, C b Represents the unit replacement cost of energy storage equipment, N r Indicates the rated number of cycles of the equipment, SoH t represents the health of the device at time t, ω t represents the market default risk weight, η t Indicates the scheduling instruction execution rate, n exec Indicates the number of historical execution instructions, n order Indicates the total number of historical instructions issued, |ΔE t | represents the absolute value of the charge and discharge energy, and T represents the number of time steps in the entire scheduling period; In step S2, the step of calculating the switching cost threshold is: Before each charge and discharge state switch, a dynamic threshold is used to measure the switching cost: Among them, Θ t represents the switching cost threshold at time t, α, β, and γ represent the weighted coefficients of loss cost, market default risk, and electricity price fluctuation range, respectively, and α+β+γ=1. represents the average electricity price within the rolling window, Indicates the deviation of the electricity price from the average electricity price.

5. The method for constructing a power transaction settlement model based on a virtual power plant according to claim 1, characterized in that: In step S3, during the dynamic adjustment of the settlement period length, the settlement period length is dynamically determined through the linkage mechanism between electricity price fluctuation rate and equipment health, and the current period is extended when the number of actions exceeds a threshold; The electricity price volatility is defined as: Among them, σ t represents the electricity price volatility at time t, W represents the time window length used to calculate the volatility, and P i represents the electricity price in period i, represents the average electricity price within the window; The normalized volatility is defined as: Among them, t represents the normalized electricity price volatility, σ ref represents the reference volatility threshold; The formula for calculating the length of the basic settlement cycle is: L t =L0(1+α3ζ t +α4(1-SoH t )), Among them, L t represents the length of the basic settlement period at time t, L0 represents the length of the benchmark period, α3 represents the influence weight of electricity price fluctuation, α4 represents the influence weight of health, and SoH t Indicates the health of the device at time t; Introducing an over-limit extension mechanism, specifically: Among them, L' t Indicates the length of the extended settlement cycle, n act Indicates the total number of charge and discharge actions within the monitoring period U, N crit It represents the critical value of the number of actions, β represents the over-limit extension coefficient, and U represents the length of the monitoring period.

6. A method for constructing a power transaction settlement model based on a virtual power plant according to claim 1, characterized in that: The sliding window filtering process in step S4 adopts a variable window width mechanism, and the window width is dynamically adjusted according to the following parameters: The current price gradient of the electricity market; The percentage of remaining available capacity of the energy storage device; The electricity consumption deviation rate between adjacent settlement cycles.

7. A method for constructing a power transaction settlement model based on a virtual power plant according to claim 6, characterized in that: In step S4, the sliding window filtering process is performed as follows: At time t, based on the price difference between the current period and the previous period, calculate the price change gradient: G t =|P t -P t-1 |, Among them, G t represents the electricity price change gradient at time t, P t represents the electricity price in period t, P t-1 represents the electricity price in period t-1; Calculate the electricity deviation rate of adjacent cycles to measure the relative difference in net electricity between two adjacent settlement cycles. The deviation rate is defined as: Among them, D t Indicates the power deviation rate between the current settlement period and the previous period. Indicates the net electricity consumption of the last settlement period. Indicates the net electricity consumption of the previous two cycles, and n indicates the sequence number of the current settlement cycle; Calculate the dynamic window width, set the base window width as W0, and combine the electricity price gradient, remaining capacity and cycle deviation rate to construct the dynamic window width: Among them, W t represents the sliding window width at time t, W0 represents the reference window width, α5 represents the influence weight of the electricity price gradient, α6 represents the influence weight of the remaining capacity, and SoC t represents the percentage of remaining available capacity at time t, α7 represents the impact weight of the cycle deviation rate, Indicates rounding down, clip(x,a,b) means limiting x to the interval [a,b], W min With W max Represents the minimum and maximum window width respectively; The high-frequency net power data x is processed by a dynamic width sliding window t Perform average filtering and output the smoothed settlement power sequence: Among them, y t represents the smoothed settlement electricity at time t, x i Represents the net power data at the i-th moment.

8. The method for constructing a power transaction settlement model based on a virtual power plant according to claim 1, characterized in that: The device health decay model in step S5 includes: Capacitance decay curve fitting module, which establishes a three-dimensional mapping relationship based on the device operating temperature, charge and discharge depth, and cycle number; The health compensation coefficient generation module converts the capacity decay rate into a settlement amount correction ratio; The compensation coefficient verification module achieves multi-party consensus confirmation through blockchain smart contracts.

9. The method for constructing a power transaction settlement model based on a virtual power plant according to claim 1, characterized in that: The specific implementation of the offline data compensation in step S6 includes: Build a user electricity consumption pattern feature vector library, including typical electricity consumption curves and abnormal pattern labels; The power consumption similarity weights between disconnected nodes and adjacent nodes are calculated based on the graph attention network; Generate compensation power estimates based on topological distance and similarity weights.

10. A method for constructing a power transaction settlement model based on a virtual power plant according to claim 9, characterized in that: The power consumption pattern feature vector library includes: Fourier transform low-frequency component coefficients of a typical daily load curve; Sliding window statistics of the proportion of peak and valley periods in electricity consumption; The marking frequency of abnormal power usage events, such as power sag and sustained overload.

Images