A method for evaluating the revenue of energy storage devices participating in power ancillary services

Abstract

This application provides a method for evaluating the revenue of energy storage devices participating in power ancillary services, including: extracting the urgency of peak-shaving services and the real-time requirements of frequency regulation response, identifying the priority ranking of each service response, allocating the mixed output data to the peak-shaving and frequency regulation parts according to priority, and obtaining the power share after preliminary decomposition; obtaining the standby loss data of standby power and matching it with the overlapping part of charging and discharging power; if the loss data matches the non-allocated residual in the power share after preliminary decomposition, the residual part is assigned to standby power, and a complete service type power allocation is obtained; extracting the revenue adjustment coefficient from the allocated revenue value, identifying the weight of the revenue adjustment coefficient among services, and reallocating if the deviation exceeds a threshold, and obtaining an optimized and accurate allocation result.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a method for evaluating the revenue of energy storage devices participating in power ancillary services. Background Technology

[0002] Energy storage devices play a crucial role in power systems, supporting the stable operation of the power grid by participating in ancillary services such as peak shaving, frequency regulation, and reserve. These services balance power supply and demand, maintain frequency stability, and ensure system reliability, playing an irreplaceable role in promoting the large-scale application of new energy sources and grid intelligence. With the deepening of power market reforms, the scenarios in which energy storage devices participate in various ancillary services are increasing. However, existing assessment methods often face challenges in handling complex scenarios, failing to accurately reflect the actual contribution of energy storage devices in different services. Currently, common methods for revenue assessment rely heavily on power allocation under a single service scenario or simple time-of-use statistics. This approach has significant shortcomings in practical operation, especially when multiple service demands occur simultaneously, making it difficult to distinguish which service the power output of the energy storage device belongs to. For example, when an energy storage device responds to both peak shaving and frequency regulation demands simultaneously, its actual discharge power may be the sum of the responses to both services, while the standby state of the reserve service introduces additional energy losses. These methods typically assume that service demands are independent and do not interfere with each other, ignoring the mutual influence when multiple services are executed in parallel, leading to a decrease in the accuracy of revenue allocation. A deeper technical challenge lies in the highly dynamic coupling of power output from energy storage devices in multi-service parallel scenarios. This dynamic coupling stems from the different priorities and time scales required by different services for the power response of energy storage devices. For example, frequency regulation services require rapid response, with frequent but small fluctuations in power output, while peak shaving services require continuous and stable power output over a longer time span. When both occur simultaneously, the actual power output of the energy storage device is difficult to clearly decompose, exhibiting a mixed state. For instance, at a certain moment, the energy storage device may simultaneously receive a frequency regulation signal requiring rapid power adjustment and a peak shaving signal requiring continuous discharge. The combined effect of these two demands means the total power output of the device cannot be directly attributed to a single service. This mixed state not only increases the complexity of revenue allocation but may also lead to the overestimation or underestimation of the contribution of some services. Therefore, accurately decomposing the mixed power output to each service type when multiple services are executed in parallel becomes a key issue in revenue assessment. Summary of the Invention

[0003] This application provides a method for evaluating the revenue of energy storage devices participating in power ancillary services, which aims to significantly improve the resource utilization efficiency of energy storage devices in complex scenarios when multiple services are executed in parallel.

[0004] This invention provides a method for evaluating the revenue of energy storage devices participating in electricity ancillary services, including:

[0005] Acquire the charging and discharging power of the energy storage device at each moment, extract the mixed output data related to peak shaving, frequency regulation, and standby, compare it with a preset benchmark threshold, identify the time period when the power output exceeds the range of a single service, extract the overlapping interval of multiple services with simultaneous power output, and mark the service overlap period; acquire the service demand signal strength and execution start and end time information at the corresponding moment, identify the priority order of scheduling instructions for each service type, and generate an active service combination list; extract the urgency of peak shaving service and the real-time requirements of frequency regulation response, identify the priority order of each service response, and, combined with the power distribution characteristics of the service overlap period, allocate the mixed output data to the peak shaving and frequency regulation parts according to priority to generate the preliminary decomposed power share; acquire the standby loss data of standby, and compare it with the... The overlapping charging and discharging power is matched, and the unallocated residual in the initial power share is assigned to the standby mode to generate a complete service type power allocation. The actual contribution of each service in the complete service type power allocation is extracted. Based on the power output duration and intensity ratio allocation rules, combined with the service superposition effect of the activated service combination list, the total revenue value is mapped to each service according to the contribution ratio to generate the allocated revenue value. The power loss overlap data and the power occupancy ratio and response time data of each service are obtained. The revenue adjustment coefficient is extracted from the allocated revenue value to generate an optimized accurate allocation result. The optimized accurate allocation result is integrated, the service type contribution and revenue value are summarized, the difference between the initial power and the separated power is verified, and the verified contribution data is generated.

[0006] Furthermore, the acquisition of the charging and discharging power of the energy storage device at each moment, extraction of mixed output data related to peak shaving, frequency regulation, and standby, comparison with a preset benchmark threshold, identification of periods where power output exceeds the range of a single service, extraction of overlapping intervals of simultaneous power output from multiple services, and marking of overlapping service periods include:

[0007] The system acquires charging and discharging power data for each sampling point of the energy storage device within the monitoring period. Instantaneous power values ​​are read and timestamps are recorded using power sensors to distinguish between charging and discharging states. The charging and discharging power data is then filtered to generate processed power time-series data. For the processed power time-series data, preset peak-shaving service power thresholds, frequency regulation service response thresholds, and standby power thresholds are read. The power values ​​at each moment are iterated, and periods where the power values ​​exceed a single service threshold are marked to generate initial overlapping periods. The mean and standard deviation of power within the initial overlapping periods are calculated to confirm multi-service parallel operation. The overlapping service periods and corresponding mixed output data are then output.

[0008] Furthermore, the step of acquiring the service demand signal strength and execution start and end time information at the corresponding time, identifying the priority order of scheduling instructions for each service type, and generating an active service combination list includes:

[0009] The system reads dispatch command data for peak shaving, frequency regulation, and standby services from the power grid dispatch center, parses the command type, target power value, frequency deviation signal value, standby power requirement, and execution duration, and generates a priority sequence arranged in the order of frequency regulation, peak shaving, and standby. It calculates the signal strength of each service in the priority sequence, marks services with signal strength exceeding a threshold and overlapping execution periods as pending activation. It checks the service combinations in the pending activation state, determines the dominant and auxiliary services based on signal strength stability within the overlapping periods, and generates a list of activated service combinations containing service type, activation start and end times, and dominant / auxiliary relationships.

[0010] Furthermore, the extraction of the urgency of peak shaving services and the real-time requirements of frequency modulation responses, the identification of the priority ranking of each service response, and the allocation of the mixed output data to the peak shaving and frequency modulation components according to priority based on the power distribution characteristics of the overlapping service periods, generate the preliminary decomposed power shares, including:

[0011] Extract the load gap and duration parameters of the peak shaving service, obtain the response speed and accuracy parameters of the frequency regulation service, and generate a priority ranking sequence for frequency regulation, peak shaving, and standby. Read the power output data during the overlapping service period, convert the time-domain power signal into a frequency-domain signal, and separate the high-frequency component corresponding to the frequency regulation service and the low-frequency component corresponding to the peak shaving service. Based on the frequency domain decomposition results, construct the state-space equation, use a filter to separate the frequency regulation power share and the peak shaving power share, adjust the allocation ratio, and output the power share after the initial decomposition.

[0012] Furthermore, the step of acquiring standby power loss data of the backup standby unit, matching it with the overlapping portion of the charging and discharging power, and assigning the unallocated residual in the preliminary power allocation to the backup standby unit to generate a complete service type power allocation includes:

[0013] Obtain the standby self-discharge and auxiliary equipment power values, read the total power time-series data of the energy storage equipment, calculate the residual power after subtracting the peak-shaving and frequency-regulation power shares from the total power; extract the average value and standard deviation of the residual power, match it with the standby loss data, and confirm that the residual power is the standby power consumption; integrate the peak-shaving, frequency-regulation and the standby power consumption, verify the difference with the total power, and output the complete service type power allocation result.

[0014] Furthermore, the extraction of the actual contribution of each service in the power allocation of the complete service type, based on the power output duration and intensity ratio allocation rules, and combined with the service superposition effect of the activated service combination list, maps the total revenue value to each service according to the contribution ratio, generating the allocated revenue value, including:

[0015] Extract the power values ​​of each service from the power allocation of the complete service type, calculate the product of peak power duration, frequency regulation amplitude, and standby power duration to generate the basic contribution; count the running time of each service in the high, medium, and low power ranges, apply weight values, calculate the weighted power duration ratio, and generate the intensity-adjusted contribution; based on the list of activated service combinations, adjust the contribution of overlapping periods, normalize the data, calculate the contribution ratio of each service, allocate the total revenue value according to the ratio, and output the allocated revenue value.

[0016] Furthermore, the acquisition of power loss overlap data and the power occupancy ratio and response time data of each service, and the extraction of revenue adjustment coefficients from the allocated revenue value to generate optimized and accurate allocation results include:

[0017] Obtain the power consumption during the execution period of each service, calculate the power occupancy ratio and response time of each service, and allocate the power consumption share; extract the net revenue of each service from the allocated revenue value, calculate the revenue adjustment coefficient, and determine the allocation weight; based on the allocation weight, calculate the corrected revenue value, iteratively adjust the weight until the deviation rate meets the threshold, and output the optimized accurate allocation result.

[0018] Furthermore, the process of integrating the optimized and precise allocation results, summarizing the service type contribution and revenue value, verifying the difference between the initial power and the power after separation, and generating verified contribution data includes:

[0019] Summarize the contribution and revenue of peak shaving, frequency regulation, and standby power to generate a data report containing timestamps, service types, and power values; calculate the difference between the original total power and the separated power in the data report, verify that the difference meets the threshold, and output the final data containing the verification identifier and contribution level.

[0020] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0021] This invention discloses a method for evaluating the revenue of energy storage devices participating in power ancillary services. Addressing the business challenges of overlapping power output, conflicting scheduling priorities, and uneven revenue distribution in multi-service scenarios such as peak shaving, frequency regulation, and standby, this method extracts actual charging and discharging power and service demand signals, identifies overlapping service periods and priority rankings, decomposes the mixed output into peak shaving, frequency regulation, and standby power shares based on response time and urgency, optimizes residual allocation by matching standby loss data, and distributes total revenue based on service contribution and response time ratios. The method dynamically adjusts allocation weights to reduce deviations, ultimately generating an accurate data report containing power allocation details and revenue distribution, and verifying the separation accuracy. This invention significantly improves the resource utilization efficiency and economic benefits of energy storage devices in complex scenarios through multi-service power decomposition and dynamic revenue distribution. Attached Figure Description

[0022] Figure 1 This is a flowchart of a method for evaluating the revenue of energy storage devices participating in power ancillary services according to the present invention.

[0023] Figure 2 This is a schematic diagram of a method for evaluating the revenue of energy storage devices participating in power ancillary services according to the present invention. Detailed Implementation

[0024] To further understand the content of this invention, a detailed description of the invention is provided in conjunction with the accompanying drawings and embodiments. The specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention. It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0025] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the embodiments of this application will be described in detail below with reference to the accompanying drawings and specific examples.

[0026] The technical solutions of the embodiments of this application will be described below with reference to the accompanying drawings. In the description of the embodiments of this application, unless otherwise stated, " / " means "or," for example, A / B can mean A or B; "and / or" in this text is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Furthermore, in the description of the embodiments of this application, "multiple" refers to two or more than two.

[0027] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this embodiment, unless otherwise stated, "a plurality of" means two or more.

[0028] Furthermore, the terms “comprising” and “having”, and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus.

[0029] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a specific manner to facilitate understanding.

[0030] It should be understood that in this application, "at least one (item)" means one or more. "More than one" means two or more. "At least two (items)" means two or three or more. "And / or" is used to describe the relationship between related objects, indicating that there can be three relationships. For example, "A and / or B" can mean: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural.

[0031] The character " / " generally indicates that the preceding and following objects are in an "or" relationship. "At least one of the following" or similar expressions refer to any combination of these items, including any single or multiple items. For example, "at least one of a, b, or c" can be expressed as: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0032] Both "...when" and "if" indicate that a corresponding action will be taken under certain objective circumstances. They are not time limits, nor do they require a judgment action to be taken when the action is taken, nor do they imply any other limitations.

[0033] The method provided in this application embodiment can be executed by an energy storage device participating in the evaluation of revenue from ancillary services. This device can be an electronic device or a device applied within an electronic device, such as an energy storage device participating in the evaluation of revenue from ancillary services. The electronic device can refer to devices such as mobile phones or servers; this application embodiment does not limit this.

[0034] like Figures 1-2 This application provides a method for evaluating the revenue of energy storage devices participating in electricity ancillary services, which may specifically include:

[0035] Step S101: Obtain the actual charging and discharging power of the energy storage device at each time, extract the mixed output data related to peak shaving, frequency regulation and standby, compare it with the preset benchmark threshold, identify the time period when the power output exceeds the range of a single service, extract the overlapping interval of power output of multiple services at the same time, and identify the overlapping service time period.

[0036] The system acquires charging and discharging power data for each sampling point of the energy storage device within the monitoring period. Instantaneous power values ​​are read using power sensors and corresponding timestamps are recorded. Charging and discharging states are distinguished based on positive and negative power values. The raw power data is filtered to remove noise interference, resulting in processed power time-series data. For the processed power time-series data, preset peak-shaving service continuous power thresholds, frequency regulation service instantaneous response thresholds, and standby minimum power thresholds are read. The power value at each moment is iterated. When the absolute power value at a certain moment is greater than the standby minimum power threshold but less than the peak-shaving service continuous power threshold, and the power change rate exceeds the frequency regulation service instantaneous response threshold, this moment is marked as a service overlap point. Consecutive service overlap points form an initial overlap period. The initial overlap period is verified by calculating the power mean and power standard deviation within the period. If the power mean is within the peak-shaving service power range and the power standard deviation exceeds a preset fluctuation threshold, it is confirmed that multiple services are running concurrently during this period. The confirmed service overlap period and the corresponding mixed output data within that period are output.

[0037] Specifically, in one implementation, power data acquisition for the energy storage device is achieved through a high-precision power sensor installed on the DC bus, with a sampling frequency set to 10 times per second to ensure the capture of rapidly changing frequency modulation response signals. The instantaneous power value output by the power sensor includes amplitude and direction information; a positive value indicates the discharge state, and a negative value indicates the charging state. A moving average filter is used for filtering, with a window width of 5 sampling points, effectively removing high-frequency noise interference while preserving the dynamic characteristics of power changes. The continuous power threshold for peak shaving service is set based on 60% to 80% of the rated power of the energy storage device, the instantaneous response threshold for frequency modulation service is set at a power change rate exceeding 5% of the rated power per second, and the minimum standby power threshold is set at 2% of the rated power. The power change rate is calculated by dividing the power difference between two adjacent sampling points by the sampling time interval. When the energy storage device outputs 30% of its rated power at a certain moment, and the power change rate reaches 8% of the rated power per second, it indicates that the device is both performing peak shaving discharge and responding to the rapid adjustment requirements of the frequency modulation signal. The marking of service overlap points is implemented using a state machine approach. At each sampling time, it is determined whether to enter an overlap state based on a combination of power value and power change rate. If more than 10 consecutive sampling points are in an overlap state, an initial overlap period is formed. In practical applications, the power grid dispatch center may simultaneously issue a peak-shaving command requiring the energy storage device to continuously discharge 500 kW, while the automatic generation control system sends a frequency regulation signal requiring rapid adjustment of ±50 kW. The actual output power of the energy storage device will fluctuate rapidly between 450 and 550 kW.

[0038] Preferably, the verification of the initial overlapping period is achieved by calculating the arithmetic mean and standard deviation of the power at all sampling points within the period. The power mean reflects the main service type of the energy storage device during that period; if the mean is within the peak-shaving service power range, it indicates that peak-shaving is the main service. The power standard deviation reflects the degree of power fluctuation; a larger standard deviation indicates frequent power regulation, which is a typical characteristic of frequency regulation services. The preset fluctuation threshold is obtained based on historical operating data and is typically set at 3% to 5% of the rated power. Through this dual verification mechanism, the periods during which the energy storage device simultaneously participates in multiple ancillary services can be accurately identified, providing a reliable data foundation for subsequent revenue sharing.

[0039] Step S102: Obtain the service demand signal strength and execution start and end time information at the corresponding time, identify the priority order of scheduling instructions for each service type, and if the signal strength exceeds the threshold and the execution time overlaps, determine that multiple services are activated at the same time, and obtain a list of activated service combinations.

[0040] Dispatch instruction data packets for peak shaving, frequency regulation, and standby services are read from the communication interface of the power grid dispatch center. The instruction type identifier, target power value, frequency deviation signal value, standby power requirement value, issuance timestamp, and estimated execution duration in the data packets are parsed. Based on the characteristics of the power grid dispatching procedures—frequency regulation having the highest response speed requirement, peak shaving having a longer duration, and standby being ready at any time—a priority sequence is generated, arranged in the order of frequency regulation, peak shaving, and standby. Signal strength is calculated for the instruction data of each service type in the priority sequence. The signal strength for peak shaving service equals the target power value divided by the rated power of the energy storage device; the signal strength for frequency regulation service equals the absolute value of the frequency deviation divided by the rated frequency deviation limit; and the signal strength for standby service equals the standby power requirement divided by the rated power. When the calculated signal strength exceeds a preset service activation threshold and there is a time overlap between the execution periods of different services, the corresponding service is marked as pending activation. Check the service combination status to be activated. If the overlapping period of the execution time of the peak shaving service and the frequency modulation service exceeds the minimum overlap duration threshold, the service with stable signal strength is determined as the dominant service and the service with fluctuating signal strength is determined as the auxiliary service based on the stability of the peak shaving signal strength and the fluctuation frequency of the frequency modulation signal strength within the overlapping period. Generate an activated service combination list containing service type identifier, activation start time, end time and dominant / auxiliary relationship marker.

[0041] Specifically, in one implementation, the power grid dispatch center sends dispatch instruction data packets to the energy storage device via the IEC 61850 standard protocol. Each data packet contains a fixed-format header and a variable-length data field. The header records the instruction sequence number, transmission timestamp, and priority identifier, while the data field carries specific service parameters. Peak-shaving service instructions include the target active power setpoint and duration requirements; frequency regulation service instructions include the frequency deviation measurement and AGC adjustment signal; and standby service instructions include the standby power level and response preparation time. The signal strength calculation reflects the urgency of various service demands. The peak-shaving service signal strength is equal to the target power value divided by the rated power of the energy storage device. When the grid requires the energy storage device to output 800 kW and the device's rated power is 1000 kW, the signal strength is 0.8. The frequency regulation service signal strength is based on the normalization of the frequency deviation. The signal strength corresponding to a frequency deviation of 0.1 Hz is the absolute value of the frequency deviation (0.1) divided by the rated frequency deviation limit of 0.2 Hz, resulting in a signal strength of 0.5. Service activation thresholds are typically set at 0.1; exceeding this threshold indicates that the service demand has reached a level requiring a response. The determination of dominant and auxiliary services is based on statistical analysis of signal characteristics. During overlapping periods, the signal strength of peak-shaving services typically remains relatively stable, with fluctuations less than 5%, while the signal strength of frequency regulation services exhibits rapid fluctuations, potentially changing multiple times per second. By calculating the coefficient of variation (COP) of the signal strength of each service during overlapping periods, services with smaller COPs are identified as dominant, and those with larger COPs as auxiliary. This method accurately identifies the primary service functions of energy storage equipment, providing a basis for revenue allocation. The list of activated service combinations is stored in a structured data format. Each record includes a service type identifier, a timestamp of the activation start time, a timestamp of the activation end time, and a primary / auxiliary relationship marker. When peak-shaving service is the dominant service, the marker value is 1; when frequency regulation service is the auxiliary service, the marker value is 2. This marker method facilitates subsequent power decomposition and revenue calculation.

[0042] In one embodiment, an energy storage power station simultaneously receives instructions for peak-shaving discharge of 500 kW and frequency regulation response of ±30 kW during the afternoon peak electricity consumption period, with the execution periods of the two services overlapping by 30 minutes. Using the above method, peak-shaving is determined to be the primary service and frequency regulation to be the secondary service. The resulting active service combination record provides an accurate identification result for the multi-service parallel status of the energy storage device during this period.

[0043] Step S103: Extract the urgency of peak shaving service and the real-time requirements of frequency regulation response, identify the priority ranking of each service response, and allocate the mixed output data to the peak shaving and frequency regulation parts according to priority to obtain the power share after preliminary decomposition.

[0044] The urgency indicators of peak-shaving services are extracted, including the grid load gap value, expected duration, and affected area parameters. Real-time requirements for frequency regulation services are obtained, including frequency deviation response speed, regulation accuracy requirements, and response delay limits. Based on the provisions in the power grid safety operation regulations that frequency regulation response time is less than a preset first threshold, peak-shaving response time is less than a preset second threshold, and reserve response time is less than a preset third threshold, a priority ranking sequence for frequency regulation, peak regulation, and reserve is generated. For the priority ranking sequence, power output data within the service overlap period is read. The time-domain power signal is converted to a frequency-domain signal using a Fast Fourier Transform. Components with frequencies higher than a preset boundary frequency are classified as high-frequency components, corresponding to the rapid power regulation characteristics of frequency regulation services. Components with frequencies lower than the preset boundary frequency are classified as low-frequency components, corresponding to the stable power output characteristics of peak-shaving services, resulting in the frequency domain decomposition of the power signal. Based on the frequency distribution characteristics in the frequency domain decomposition results, a state-space equation for the mixed power signal is constructed. A Kalman filter is used for state estimation and prediction. According to the high priority attribute of the frequency modulation service, the power component that meets the frequency regulation requirements is separated from the mixed signal as the frequency modulation power share. The remaining power component is allocated to the peak shaving service as the peak shaving power share. If the peak shaving power share is lower than the minimum requirement value of the peak shaving service, a portion of the power is transferred from the frequency modulation power share to the peak shaving power share proportionally. The matching degree of the frequency modulation power share and the peak shaving power share is verified, and the deviation rate between each share and the original service instruction is calculated. If the peak shaving power deviation rate exceeds a preset deviation threshold or the frequency modulation response delay exceeds a preset delay threshold, the power allocation ratio of the two types of services is iteratively adjusted until the performance indicators of each service meet the preset requirements. The preliminary decomposed peak shaving power share and frequency modulation power share are then output.

[0045] Specifically, in one implementation, the urgency index of peak-shaving services is obtained by real-time monitoring of the deviation between the grid load curve and the generation plan. The grid load gap value equals the actual load demand minus the available generation capacity. When the gap value exceeds 50% of the system's reserve capacity, the urgency level is increased. The expected duration value is calculated based on the changing trend of the load forecast curve, and the duration of high load is estimated by analyzing historical load patterns. The affected area parameter is determined by statistically analyzing the number of affected substations and the number of power users; the more critical users covered, the higher the urgency level. The application of Fast Fourier Transform (FFT) in power signal decomposition is based on the principle of frequency domain analysis. The mixed power output signal of energy storage devices contains different frequency components. The power change corresponding to peak-shaving services is slow, mainly concentrated in the low-frequency band of 0.001 Hz to 0.01 Hz, while the power adjustment of frequency regulation services is rapid, with frequency components distributed in the range of 0.1 Hz to 1 Hz. By performing an FFT transform on the power time series within the sampling period, the spectral distribution of the power signal is obtained. A boundary frequency of 0.05 Hz is set. The amplitudes of spectral lines below this frequency are accumulated to obtain the low-frequency power component, while the amplitudes of spectral lines above this frequency are accumulated to obtain the high-frequency power component. This frequency domain decomposition method can effectively distinguish service response characteristics at different time scales. Constructing the state-space equation is a key step in achieving accurate power allocation. The state vector of the mixed power signal contains three state variables: the current power value, the rate of change of power, and the cumulative power. The observation vector is the actual measured power output value. The state transition matrix is ​​determined based on the response characteristics of the energy storage device, reflecting the evolution of power from the current state to the next state. The process noise covariance matrix characterizes the uncertainty of the system, while the observation noise covariance matrix reflects the statistical characteristics of the measurement error. The Kalman filter estimates the optimal state of the system at each sampling time through two recursive steps: prediction and update.

[0046] x k+1 This represents the state vector at the next moment, containing three state variables: power value, rate of change of power, and cumulative power. A k The state transition matrix represents the response characteristics and power evolution of the energy storage device. k w represents the state vector at the current moment. k This represents the process noise vector, characterizing the system uncertainty.

[0047] , This represents the prior state estimate of the current time step based on information from the previous time step, where Fk represents the state transition matrix used to predict the power state evolution. Let Bk represent the posterior state estimate of the previous time step, and let u represent the control input matrix. k This represents the control input vector.

[0048]

[0049] , This represents the posterior state estimate at the current moment, i.e., the corrected optimal power state. K represents the prior state estimate. k Let z represent the Kalman gain matrix. k H represents the observation vector at the current moment, i.e., the actual measured power value. k The observation matrix represents the mapping from the state space to the observation space. The prediction step calculates the prior estimate for the current time step based on the state estimate from the previous time step and the state transition equation. The update step corrects the prior estimate by combining the observations from the current time step to obtain the posterior state estimate.

[0050] Preferably, power allocation follows a priority principle and capacity constraints. Frequency modulation (FM) services have the highest priority, and their power requirements are initially extracted from the mixed signal. Based on real-time changes in the frequency deviation signal, the required power adjustment for FM is calculated, and this adjustment is directly separated from the total power. Remaining power is allocated to peak shaving services, but must meet their minimum power requirements. When remaining power is insufficient, the sensitivity coefficient of the FM response is reduced to decrease FM power occupancy, freeing up some power for peak shaving services.

[0051] In one possible implementation, matching verification uses a sliding window approach. The time window length is set to 1 minute, and the average and standard deviation of each service power share are calculated within the window. The peak-shaving power deviation rate is equal to the difference between the actual allocated power and the commanded power divided by the commanded power, with a deviation threshold set at ±10%. The frequency modulation response delay is calculated using a cross-correlation function to determine the time lag between the frequency deviation signal and the power response signal, with a delay threshold set at 200 milliseconds.

[0052] For example, the iterative adjustment process uses gradient descent to optimize the allocation ratio. The objective function is defined as the weighted sum of squares of the deviation rates of each service, and the weight coefficients are set according to the service priority. In each iteration, the gradient of the objective function with respect to the allocation ratio is calculated, and the allocation ratio is adjusted in the opposite direction of the gradient, with the step size dynamically adjusted according to the convergence speed. The iteration terminates when the change in the objective function is less than a preset convergence threshold after three consecutive iterations.

[0053] Understandably, a certain energy storage power station responds to both peak shaving and frequency regulation demands during the evening peak hours, with a total output power of 800 kW. 100 kW of frequency regulation power is allocated for rapid frequency adjustment, and 700 kW of peak shaving power is allocated for continuous power support. This allocation scheme satisfies both the rapid response requirement for grid frequency stability and the power support effect of peak shaving and valley filling. Furthermore, the initially decomposed power share data is stored in time-series format, with each data point containing a timestamp, peak shaving power value, frequency regulation power value, and allocation quality identifier. The allocation quality identifier is used to indicate whether the allocation result at that moment meets performance requirements, providing data support for subsequent revenue calculation and allocation optimization.

[0054] Step S104: Obtain the standby loss data of the standby device and match it with the overlapping part of the charging and discharging power. If the loss data matches the unallocated residual in the power share after preliminary decomposition, the residual part is assigned to the standby device to obtain the complete service type power allocation.

[0055] The system acquires standby power consumption data for standby services, including the self-discharge power value of the energy storage device in standby mode and the operating power value of auxiliary equipment. It reads the time-series data of the total power output of the energy storage device and extracts the peak-shaving power share and frequency-modulation power share at each moment from the peak-shaving and frequency-modulation power decomposition results. It then calculates the residual power value obtained by subtracting the peak-shaving and frequency-modulation shares from the total power. Feature extraction is performed on the residual power value, and the average value and standard deviation of the residual power are calculated. These are then compared with the self-discharge power and auxiliary equipment power in the standby power consumption data. If the deviation between the average residual power value and the standby power consumption is within a preset tolerance range and the standard deviation of the residual power is less than a preset fluctuation threshold, the residual power is confirmed as the power consumption for standby services. Based on the confirmed standby power consumption, it is assigned to the standby service type. The system integrates the peak-shaving power share, frequency-modulation power share, and standby power consumption. The power of the three service types is accumulated and the difference is calculated with the original total power. If the absolute value of the difference is less than a preset error threshold, a complete service type power allocation result including peak-shaving, frequency-modulation, and standby power is output.

[0056] Specifically, in one implementation, the standby loss of the backup standby service mainly comprises three components. Self-discharge power is the energy loss of the energy storage battery due to internal chemical reactions during standby, typically 0.1% to 0.3% of its rated capacity per day. Auxiliary equipment operating power includes the continuous power consumption of the battery management system, communication modules, and monitoring equipment, which remain operational during standby to maintain the basic functions of the energy storage system.

[0057] Specifically, the feature extraction and matching process for residual power needs to comprehensively consider the stability and amplitude characteristics of the power. After obtaining the total power of the energy storage device, subtracting the identified peak-shaving and frequency-regulating power shares, the resulting residual power usually exhibits relatively stable characteristics. By calculating the average value of the residual power over a continuous period and comparing it with a pre-determined standby loss benchmark value, when the deviation between the two is within ±5%, the residual power can be preliminarily determined to belong to standby loss. The standard deviation is used to assess the degree of fluctuation of residual power. The power fluctuation in standby mode should be significantly less than the power fluctuation of peak-shaving or frequency-regulating services, and the standard deviation is usually no more than 10% of the average value. The determination of the preset tolerance range and fluctuation threshold is based on the statistical analysis of historical operating data of the energy storage device. By analyzing the power consumption characteristics in pure standby mode, a feature template for standby loss is established. The tolerance range considers the influence of factors such as ambient temperature, battery aging degree, and auxiliary equipment operating status, while the fluctuation threshold reflects the stability of power consumption in standby mode.

[0058] Preferably, integrity verification is achieved through power balance testing. The identified peak-shaving power share, frequency-modulated power share, and standby power consumption are summed, and the difference between this sum and the original total power is calculated. A preset error threshold is typically set at 1% of the total power. When the absolute value of the difference is less than this threshold, it indicates that the power allocation result satisfies the principle of energy conservation, and the power allocation for each service type is accurate.

[0059] In one embodiment, during the low-load period at night, a certain energy storage power station has a total output power of 50 kW, of which the peak-shaving power share is 0 kW, the frequency regulation power share is 45 kW, and the remaining 5 kW of residual power matches the standby loss characteristics and is accurately attributed to the power consumption of the backup standby service.

[0060] Step S105: Extract the actual contribution of each service in the power allocation of the complete service type, identify the proportional allocation pattern based on the power output duration and intensity, combine the service superposition effect of the activated service combination list, and map the total revenue value to each service according to the contribution ratio to obtain the allocated revenue value.

[0061] The power values ​​of each service in each time period are extracted from the power allocation results of the complete service type. The sum of the products of the power value and execution duration of the peak-shaving service, the sum of the products of the number of power adjustments and the absolute value of each adjustment amplitude of the frequency-adjustment service, and the product of the standby duration and average standby power of the backup standby service are calculated. The basic contribution is obtained based on the ratio of the power duration product of each service to the total power duration product of all services. For this basic contribution, the runtime of each service at different power levels is statistically analyzed. Power levels are divided into high, medium, and low ranges according to the percentage of rated power. The runtime in the high-power range is multiplied by a preset high weight value, the runtime in the medium-power range by a preset medium weight value, and the runtime in the low-power range by a preset low weight value. These are accumulated to obtain the weighted power duration. The proportion of the weighted power duration of each service to the total weighted power duration is calculated to obtain the contribution after intensity adjustment. Based on the adjusted contribution amount, the time periods during which multiple services are executed simultaneously are identified. When peak shaving and frequency regulation are executed simultaneously, the peak shaving contribution amount is multiplied by a preset peak shaving superposition coefficient, and the frequency regulation contribution amount is multiplied by a preset frequency regulation superposition coefficient. The contribution amount for each individual execution period remains unchanged. The adjusted contribution amounts for each time period are accumulated to obtain the corrected contribution amount considering the superposition effect. The corrected contribution amount is normalized, and the percentage of the corrected contribution amount of each service (peak shaving, frequency regulation, and standby) to the total corrected contribution amount of the three types of services is calculated. The total revenue value of the energy storage equipment during the evaluation period is obtained. The total revenue value is multiplied by the percentage of each service to obtain the apportioned revenue value of each service (peak shaving, frequency regulation, and standby).

[0062] Specifically, in one implementation, the calculation of the basic contribution requires differentiated quantification methods based on the characteristics of different service types. The contribution of peak-shaving service is mainly reflected in continuous and stable power output, and its contribution is measured by accumulating the power value of each time period and multiplying it by the corresponding duration. The contribution of frequency regulation service is reflected in frequent and rapid power adjustment response, requiring the statistical analysis of the number of adjustments per unit time and the absolute value of each adjustment amplitude. Although the standby service has a smaller power output, its continuous online characteristic ensures system reliability, and its contribution is quantified by multiplying the standby duration by the average power.

[0063] Specifically, the power level classification is based on the operating characteristics of energy storage devices and grid demand characteristics. The high-power range is defined as above 70% of rated power; operation in this range indicates that energy storage devices are undertaking important grid support tasks, and the corresponding weight value is set at 1.5. The medium-power range is 30% to 70% of rated power, representing normal operation, and the weight value is 1.0. The low-power range is below 30% of rated power, mainly corresponding to ancillary services or standby status, and the weight value is 0.5. This differentiated weighting reflects the actual value difference in the contribution of energy storage devices to the grid at different power levels. By statistically analyzing the operating time of each service at different power levels and multiplying it by the corresponding weight coefficient, the actual contribution intensity of each service can be more accurately reflected. The specific values ​​of the weight coefficients can be dynamically adjusted according to electricity market price signals and grid operation demands, achieving market-based and refined contribution assessment.

[0064] It should be noted that handling the superposition effect is the core challenge in revenue sharing under multi-service parallel scenarios. When peak shaving and frequency regulation services are executed simultaneously, the power output of energy storage devices exhibits composite characteristics, and simple linear superposition cannot accurately reflect the actual contribution of each service. The peak shaving superposition coefficient is usually set between 0.9 and 1.1, with the upper limit taken when peak shaving is the dominant service and the lower limit taken when it is an auxiliary service. The setting of the frequency regulation superposition coefficient is more complex, requiring consideration of the response speed and accuracy requirements of frequency regulation, and is generally set between 0.8 and 1.2. The specific value of the coefficient is determined by analyzing the actual effects of multi-service parallel periods in historical operating data and optimized based on feedback from grid dispatch.

[0065] Preferably, the calculation of the corrected contribution amount adopts a time-segmented cumulative method. Within each evaluation period, it is determined whether the service is executed individually or cumulatively based on its execution status. The contribution amount for individual execution periods is directly accumulated, while the contribution amount for cumulative execution periods needs to be corrected by applying an appropriate cumulative coefficient. This approach ensures the accuracy of the contribution amount calculation while avoiding the problem of inflated contribution amounts caused by duplicate calculations.

[0066] In one possible implementation, normalization ensures the rationality and fairness of revenue distribution. By calculating the percentage of each service's contribution to the total contribution, a mapping relationship between contribution and revenue is established. The total revenue typically includes multiple components such as capacity compensation revenue, electricity revenue, and ancillary service compensation revenue, which are allocated according to the normalized percentages to ensure that each service receives an economic return commensurate with its contribution.

[0067] For example, a certain energy storage power station has a total revenue of 100,000 yuan in a settlement cycle. Calculations using the above method show that the contribution percentages for peak shaving services are 45%, frequency regulation services 40%, and standby services 15%. Accordingly, peak shaving services are allocated 45,000 yuan, frequency regulation services 40,000 yuan, and standby services 15,000 yuan. This revenue allocation mechanism based on actual contribution incentivizes energy storage equipment operators to optimize multi-service response strategies and improve equipment utilization efficiency. Furthermore, if the response quality of a certain service does not meet the grid's requirements, its allocated revenue needs to be reduced accordingly. By establishing a linkage mechanism between service quality and revenue, the system promotes the maintenance of response quality for various services when multiple services are running concurrently, achieving a balance between economic benefits and technical performance.

[0068] Understandably, this revenue-sharing method based on contribution provides a fair and transparent settlement basis for energy storage devices to participate in the electricity market.

[0069] Step S106: Extract the revenue adjustment coefficient from the allocated revenue value, identify the weight of the revenue adjustment coefficient among services, and reallocate if the deviation exceeds the threshold to obtain the optimized and accurate allocation result.

[0070] The system acquires overlapping power loss data across service execution periods, including power loss during peak shaving services, additional losses during frequency modulation service responses, and basic losses during standby. It calculates the power occupancy ratio of each service to the total power and the actual response time of each service. Power loss is then allocated according to the power occupancy ratio of each service to obtain the share of power loss each service should bear. Revenue data for each service is extracted from the initial allocated revenue value. The net revenue of each service after deducting losses is calculated based on the loss share. The ratio of net revenue to initial allocated revenue is used as a revenue adjustment coefficient. The allocation weight of the adjustment coefficient among services is determined based on the power occupancy ratio and response time ratio of each service. Based on the allocation weight and revenue adjustment coefficient, the corrected revenue value for each service is calculated. The initial allocated revenue is multiplied by the corresponding adjustment coefficient to obtain the adjusted revenue. The deviation rate between the adjusted revenue and the initial allocated revenue is calculated. If the deviation rate of any service exceeds a preset threshold, the allocation weight is readjusted proportionally to the reverse of the deviation rate. The readjusted allocation weights are iteratively corrected. In each iteration, the weight values ​​are adjusted according to the deviation rate to reduce the weight deviation of services with excessive deviation rates until the revenue deviation rate of all services is less than the preset threshold. The optimized and accurate allocation results of each service are then output.

[0071] Specifically, in one implementation, the acquisition and allocation of power losses need to consider the energy consumption characteristics of energy storage devices under different service states. Losses during peak-shaving services mainly originate from internal resistance losses and conversion efficiency losses during high-power charging and discharging processes; these losses are proportional to the square of the power. Additional losses during frequency regulation service responses stem from frequent power direction switching and efficiency reductions caused by rapid responses; each power adjustment generates transient losses. Basic losses in standby mode include battery self-discharge, energy consumption of the temperature control system, and power consumption of monitoring equipment; these losses are relatively stable and continuous. The allocation of loss shares is primarily based on power occupancy ratios. If, during a certain period, peak-shaving service occupies 600 kW, frequency regulation service occupies 100 kW, and standby mode occupies 50 kW, for a total power of 750 kW, then the power occupancy ratios for peak-shaving, frequency regulation, and standby are 80%, 13.3%, and 6.7%, respectively. The total losses generated during this period are allocated to each service according to this ratio, reflecting the principle of "whoever uses it, bears the cost." Response time is used as an auxiliary factor to adjust the loss sharing ratio between long-term low-power services and short-term high-power services, so as to make the allocation result more reasonable.

[0072] It's important to note that the calculation of the revenue adjustment factor involves introducing the concept of net revenue. Net revenue equals the initial allocated revenue minus the attributable loss costs, reflecting the actual economic value of each service after deducting losses. The adjustment factor is derived from the ratio of net revenue to initial allocated revenue. This factor, ranging from 0 to 1, characterizes the degree of impact of losses on the revenue of each service. The closer the adjustment factor is to 1, the smaller the proportion of losses for that service, and the better its economic efficiency. The determination of the allocation weights comprehensively considers the power occupancy ratio and the response time ratio, using a weighted average method. The weighting factor for the power occupancy ratio is typically set to 0.7, and the weighting factor for the response time ratio is set to 0.3. This weighting setting reflects the dominant role of power occupancy in loss generation while also considering the impact of service duration.

[0073] Preferably, the calculation of the adjusted revenue value and the deviation control mechanism ensure the rationality of the allocation result. The initial allocated revenue is multiplied by the corresponding adjustment coefficient to obtain the adjusted revenue considering the impact of losses. The deviation rate is calculated by dividing the difference between the adjusted revenue and the initial allocated revenue by the initial allocated revenue, reflecting the degree of impact of losses on revenue. A preset threshold is typically set at 15%. When the deviation rate of a service exceeds this threshold, it indicates that the service is bearing excessive loss costs and a weight adjustment is required.

[0074] In one possible implementation, weight readjustment follows a reverse proportional allocation principle. Services with higher deviation rates should have their assigned weights reduced, thereby decreasing their share of losses. The specific adjustment amount is proportional to the deviation rate; for every 1 percentage point the deviation rate exceeds a threshold, the weight is adjusted by 0.5 percentage points. This gradual adjustment avoids drastic changes in weights, ensuring system stability.

[0075] For example, the iterative correction process gradually optimizes the allocation results through multiple rounds of adjustments. In each iteration, the revenue deviation rate of each service under the current weight is calculated first, services with deviation rates exceeding the standard are identified, and adjustment priorities are determined according to the magnitude of the deviation rate. The weight adjustment adopts a small-step strategy, with each adjustment not exceeding 5% of the original weight to prevent oscillations caused by over-adjustment. The iteration termination condition is set when the deviation rate of all services is less than a preset threshold, or the weight change in three consecutive iterations is less than 0.1%, indicating that the system has converged to a stable state. Furthermore, the weight deviation of services with deviation rates exceeding the standard is obtained by calculating the difference between the actual weight and the ideal weight. The ideal weight is the weight value that makes the deviation rate of the service exactly equal to the threshold, which can be obtained by reverse calculation. The process of reducing the weight deviation is actually a process of approaching the ideal weight, and each iteration reduces the gap between the actual weight and the ideal weight. After four rounds of iterative optimization, the final allocated revenue of a certain energy storage power station is 42,000 yuan for peak shaving services, 38,000 yuan for frequency regulation services, and 20,000 yuan for standby services. Compared to the initial allocation, the revenue from peak shaving services has decreased slightly, while the revenue from frequency regulation and standby services has increased. This adjustment reflects the differences in the actual loss characteristics of each service and achieves a more accurate and equitable distribution of revenue.

[0076] Step S107: Integrate the optimized and accurate allocation results, summarize the actual contribution and revenue of service types, form a final data report containing power allocation details and revenue allocation results, evaluate the difference between the initial power and the power after separation, verify the separation accuracy, and obtain the verified contribution data.

[0077] The optimized and precise allocation results are integrated, and the contribution and revenue values ​​of each service—peak shaving, frequency regulation, and standby—are summarized. Power data is arranged by time series, and revenue data is categorized by service type, forming a data report containing timestamps, service types, power values, contribution amounts, and allocated revenue. The original total power of the energy storage device and the power of each service separated in the data report are obtained. The difference between the original total power and the sum of the power of each service is calculated. If the absolute value of the difference is less than a preset accuracy threshold, the verification is successful, and the final data containing a verification identifier and the contribution of each service is output.

[0078] Specifically, in one implementation, the data reports are generated using a structured data format. Each record includes a timestamp of the evaluation time, a service type identifier, the actual power value, the contribution amount, and the allocated revenue amount. Peak-shaving service records are identified as 1, frequency regulation services as 2, and standby services as 3. The timestamp accuracy reaches the second level, ensuring that details of power changes are recorded.

[0079] Specifically, the verification process is achieved through the power balance principle. The original total power is collected from the total output of the energy storage device, and the power of each service is extracted from the data report and accumulated. The accuracy threshold is set at 2% of the total power. When the absolute value of the difference is less than this threshold, it indicates that the power separation process satisfies the law of conservation of energy, and the verification is successful.

[0080] Preferably, the final data includes a verification identifier field, where a value of 1 indicates successful verification and a value of 0 indicates that re-verification is required. The contribution of each service is expressed as a percentage, making it easy to intuitively understand the role of each service in the overall operation.

[0081] For example, during a certain assessment period, the peak shaving contribution was 45%, the frequency regulation contribution was 38%, and the standby contribution was 17%. These data provide accurate basis for the operation decision-making and market settlement of energy storage equipment.

[0082] The above-disclosed embodiments are merely preferred embodiments of the present invention and should not be construed as limiting the scope of the invention. Those skilled in the art will understand that implementing all or part of the above-described embodiments and making equivalent changes in accordance with the claims of the present invention are still within the scope of the invention.

Claims

1. A method for evaluating the revenue of energy storage devices participating in electricity ancillary services, characterized in that, include: Extract mixed output data related to multiple services from the charging and discharging power of the energy storage device at different times, and identify the service overlap period when multiple services are simultaneously served. The service overlap period is the overlapping interval of multiple services with simultaneous power output during the period when the power output exceeds the range of a single service. The multiple services include at least peak shaving, frequency regulation and standby. Based on the power distribution characteristics of the overlapping service periods, the mixed output data is allocated to the peak shaving and frequency modulation parts according to priority to generate a preliminary decomposed power share. Before allocating the mixed output data to the peak shaving and frequency modulation parts according to priority, the urgency of the peak shaving service and the real-time response requirements of the frequency modulation are extracted, and the priority ranking of each service response is identified. Obtain the standby power loss data of the standby device, match it with the overlapping part of the charging and discharging power, assign the unallocated residual in the preliminary decomposed power share to the standby device, and generate a complete service type power allocation; Extract the actual contribution of each service in the power allocation of the complete service type. Based on the power output duration and intensity ratio allocation rules, and combined with the service superposition effect of the activated service combination list, map the total revenue value to each service according to the contribution ratio to generate the allocated revenue value. The service superposition effect of the activated service combination list refers to: extracting the power value of each service in each time period from the power allocation result of the complete service type, calculating the sum of the products of the power value and execution duration of the peak-shaving service, the sum of the products of the number of power adjustments and the absolute value of each adjustment amplitude of the frequency-modulation service, and the product of the standby duration and average standby power of the standby service. The basic contribution is obtained according to the ratio of the power duration product of each service to the sum of the power duration products of all services. The basic contribution is calculated by taking the proportion of the weighted power duration of each service to the total weighted power duration to obtain the intensity-adjusted contribution. Based on the intensity-adjusted contribution, the time periods in which multiple services are executed simultaneously are identified. When peak shaving and frequency modulation are executed simultaneously, the peak shaving contribution is multiplied by a preset peak shaving superposition coefficient, and the frequency modulation contribution is multiplied by a preset frequency modulation superposition coefficient. The contribution of each execution period remains unchanged. The adjusted contribution of each time period is accumulated to obtain the corrected contribution considering the superposition effect. In this process, the service demand signal strength and execution start and end time information at the corresponding time are obtained, the priority order of scheduling instructions for each service type is identified, and if the signal strength exceeds the threshold and the execution time overlaps, it is determined that multiple services are activated simultaneously, and a list of activated service combinations is obtained. Acquire power loss overlap data and power occupancy ratio and response time data of each service. Extract revenue adjustment coefficients from the allocated revenue value to generate optimized accurate allocation results. The allocation weight of the adjustment coefficients among services is determined based on the power occupancy ratio and response time ratio of each service. Based on the allocation weights and revenue adjustment coefficients, calculate the corrected revenue value of each service. Multiply the initial allocated revenue by the corresponding adjustment coefficient to obtain the adjusted revenue. Calculate the deviation rate between the adjusted revenue and the initial allocated revenue. If the deviation rate of any service exceeds a preset threshold, readjust the allocation weights according to the reverse proportion of the deviation rate. The re-adjusted allocation weights are iteratively corrected. In each iteration, the weight values ​​are adjusted according to the deviation rate until the revenue deviation rate of all services is less than the preset threshold. The optimized and accurate allocation results of each service are then output. By integrating the optimized and accurate allocation results, summarizing the contribution and revenue of each service type, verifying the difference between the original total power of the energy storage device and the power after the separation of each service, and generating verified contribution data.

2. The method for evaluating the revenue of energy storage devices participating in power ancillary services according to claim 1, characterized in that, The extraction of mixed output data related to multiple services from the charging and discharging power of the energy storage device at different times, and the identification of service overlap periods where multiple services are simultaneously served, includes: Acquire the charging and discharging power data of the energy storage device at each sampling point during the monitoring period; The instantaneous power value is read by a power sensor and a timestamp is recorded to distinguish between charging and discharging states. The charging and discharging power data is then filtered to generate processed power time-series data. For the processed power time-series data, the preset peak-shaving service power threshold, frequency modulation service response threshold, and standby power threshold are read, the power value at each moment is traversed, the time period when the power value exceeds a single service threshold is marked, and the initial overlapping time period is generated. Calculate the power mean and standard deviation within the initial overlapping period, confirm the parallel operation of multiple services, and output the service overlapping period and corresponding mixed output data.

3. The method for evaluating the revenue of energy storage devices participating in power ancillary services according to claim 1, characterized in that, The process involves acquiring the service demand signal strength and execution start and end time information at the corresponding time, identifying the priority order of scheduling instructions for each service type, and determining that multiple services are simultaneously active if the signal strength exceeds a threshold and the execution times overlap. This results in a list of active service combinations, including: Read dispatch command data for peak shaving, frequency regulation, and reserve services from the power grid dispatch center, parse the command type, target power value, frequency deviation signal value, standby power requirement, and execution duration, and generate a priority sequence arranged in the order of frequency regulation, peak shaving, and reserve. Calculate the signal strength of each service in the priority sequence, and mark services whose signal strength exceeds the threshold and whose execution time overlap as pending activation. The service combination to be activated is examined. Based on the stability of the peak-shaving signal strength and the fluctuation frequency of the frequency-modulated signal strength during the overlapping period, the dominant service and the auxiliary service are determined, and the list of activated service combinations is generated. The list of activated service combinations includes the service type, activation start and end time, and dominant-auxiliary relationship.

4. The method for evaluating the revenue of energy storage devices participating in power ancillary services according to claim 1, characterized in that, The extraction of the urgency of peak shaving services and the real-time requirements of frequency modulation response, and the identification of the priority ranking of each service response, include: Extract the load gap and duration parameters of peak shaving service, obtain the response speed and accuracy parameters of frequency regulation service, and generate a priority sorting sequence for frequency regulation, peak shaving, and standby.

5. The method for evaluating the revenue of energy storage devices participating in power ancillary services according to claim 1, characterized in that, The process involves combining the power distribution characteristics of the overlapping service periods to allocate the mixed output data to the peak-shaving and frequency-modulating components according to priority, generating a preliminary power share breakdown, including: Read the power output data during the service overlap period, convert the power output data from time domain power signal to frequency domain signal, classify the components with frequencies higher than the preset threshold frequency as high frequency components, corresponding to the fast power adjustment characteristics of frequency modulation service, and classify the components with frequencies lower than the preset threshold frequency as low frequency components, corresponding to the stable power output characteristics of peak shaving service. Based on the frequency domain decomposition results, a state-space equation is constructed, and a filter is used to separate the frequency modulation power share and the peak modulation power share. The allocation ratio is adjusted, and the power share after the initial decomposition is output.

6. The method for evaluating the revenue of energy storage devices participating in power ancillary services according to claim 1, characterized in that, The process of acquiring the standby power loss data of the backup standby unit, matching it with the overlapping portion of the charging and discharging power, and assigning the unallocated residual in the preliminary power allocation to the backup standby unit to generate a complete service type power allocation includes: Obtain the standby self-discharge and auxiliary equipment power values, read the total power time-series data of the energy storage equipment, and calculate the residual power after subtracting the peak-shaving and frequency-modulation power shares from the total power. The average value and standard deviation of the residual power are extracted and matched with the standby power consumption data to confirm that the residual power is the standby power consumption. Integrate the peak-shaving power share, frequency-modulated power share, and the standby power consumption. Calculate the difference between the sum of the three service types and the original total power. If the absolute value of the difference is less than a preset error threshold, output the complete service type power allocation result.

7. The method for evaluating the revenue of energy storage devices participating in power ancillary services according to claim 1, characterized in that, The step of extracting the actual contribution of each service in the power allocation of the complete service type, based on the power output duration and intensity ratio allocation rules, and combined with the service superposition effect of the activated service combination list, maps the total revenue value to each service according to the contribution ratio, generating the allocated revenue value, including: Extract the power values ​​of each service from the power allocation of the complete service type, calculate the product of peak shaving power duration, frequency modulation adjustment amplitude, and standby power duration to generate a basic contribution; statistically analyze the running time of each service in the high, medium, and low power ranges, apply weight values, calculate the weighted power duration ratio, and generate the intensity-adjusted contribution; based on the active service combination list and the intensity-adjusted contribution, identify the time periods when multiple services are executed simultaneously. When peak shaving and frequency modulation are executed simultaneously, the peak shaving contribution is multiplied by a preset peak shaving superposition coefficient, the frequency modulation contribution is multiplied by a preset frequency modulation superposition coefficient, and the contribution of each individual execution period remains unchanged. Accumulate the adjusted contribution of each time period to obtain the corrected contribution considering the service superposition effect; normalize the corrected contribution, calculate the contribution ratio of each service, allocate the total revenue value according to the ratio, and output the allocated revenue value.

8. The method for evaluating the revenue of energy storage devices participating in power ancillary services according to claim 1, characterized in that, The process of acquiring overlapping power loss data, power occupancy ratios of each service, and response time data, extracting revenue adjustment coefficients from the allocated revenue value, and generating optimized, precise allocation results includes: Obtain the power consumption during the execution period of each service, calculate the power occupancy ratio and response time of each service, and allocate the power consumption share; determine the allocation weight of the adjustment coefficient among the services based on the power occupancy ratio and response time ratio of each service; calculate the corrected revenue value of each service based on the allocation weight and revenue adjustment coefficient, iteratively adjust the weight until the deviation rate meets the threshold, and output the optimized accurate allocation result.

9. The method for evaluating the revenue of energy storage devices participating in power ancillary services according to claim 1, characterized in that, The process integrates the optimized and precise allocation results, summarizes the service type contribution and revenue value, verifies the difference between the initial power and the power after separation, and generates verified contribution data, including: Summarize the contribution and benefit values ​​of peak shaving, frequency regulation, and standby power, and generate a data report containing timestamps, service types, and power values; Calculate the difference between the original total power of the energy storage device and the power of each service separated in the data report; If the difference is less than a preset accuracy threshold, the final data containing the verification identifier and the contribution of each service is output, and the verification identifier indicates that the verification was successful.

Images