User reliability responsibility measuring and calculating method and system considering flexibility constraint and fairness optimization

Abstract

The invention provides a user reliability responsibility measuring and calculating method and system considering flexibility constraint and fairness optimization, and the method comprises the steps: firstly obtaining the intra-day data of each user of a power system, forming a net load curve, and obtaining a thermal power generating unit scheduling plan and the maximum callable capacity of each thermal power generating unit in each time period; calculating a power shortage expected value of the system in each time period; calculating a power shortage expected value of the power system in each time period under each user absence scene; constructing a linear programming model based on a minimum kernel; and finally, obtaining the time sequence final reliability responsibility of each user in each time period. According to the method, a complete method system of time sequence risk accurate assessment-total responsibility fair responsibility determination-time sequence characteristic weight mapping is constructed, the fairness and stability of user reliability responsibility allocation are effectively improved, and meanwhile, through marginal responsibility-based time sequence allocation, the incentive effect of guiding users to optimize power consumption behaviors by responsibility signals is achieved.

Description

Technical Field

[0001] This invention relates to the field of power technology, and in particular to a method and system for calculating user reliability responsibility that takes into account flexibility constraints and fairness optimization. Background Technology

[0002] Driven by the "dual carbon" goal, building a new power system with new energy sources as the mainstay has become the core path for energy transformation. With the integration of high-proportion renewable energy sources such as wind and solar power, the power system's power structure and operating characteristics are undergoing profound changes. Existing power system reliability assessment and planning methodologies face multiple challenges in adapting to this transformation. First, there is a discrepancy between traditional generator unit available capacity modeling and system functional positioning. Traditional reliability assessments often focus on medium- to long-term time scales such as annual or monthly assessments, emphasizing the macro-balance between system static adequacy and predicted load. Thermal power units are typically simplified to a two-state probabilistic model of "full generation - outage." This modeling approach ignores the flexibility and regulation constraints of thermal power units and is only applicable to the traditional scenario of thermal power as a baseload power source. It is difficult to characterize the "actual callable capacity" limited by flexibility constraints after thermal power transforms into a "regulatory power source" in the new power system. Second, there is a lack of scientific mechanisms for tracing system reliability risks and allocating responsibility. As the cost of ensuring system reliability continues to rise, how to fairly and reasonably allocate this cost has become a core challenge in power market design. Existing cost-sharing methods are mostly based on electricity consumption or peak load ratios, lacking precise quantification of the causal relationship between user electricity consumption behavior and system time-series reliability risks. Without the guidance of the "cause-pays" principle, it is difficult to form an effective responsibility signal to guide users to smooth peak and valley loads, potentially leading to unreasonable cross-subsidies and "free-riding" phenomena. Therefore, there is an urgent need to construct a system time-series reliability assessment method that can deeply couple system flexibility constraints with real-time performance. Based on this, by accurately tracing the source and fairly and effectively solving the problem of user responsibility allocation for system reliability, a scientific and fair quantitative basis and theoretical support can be provided for establishing a new incentive-compatible power system reliability assurance mechanism. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide a method and system for calculating user reliability responsibility that takes into account flexibility constraints and fairness optimization, effectively improving the fairness and stability of user reliability responsibility allocation, and at the same time, realizing the incentive effect of responsibility signal guiding users to optimize electricity consumption behavior through time-sequential allocation based on marginal responsibility.

[0004] To achieve the above objectives, the present invention adopts the following technical solution: a method for calculating user reliability responsibility that takes into account flexibility constraints and fairness optimization, comprising the following steps:

[0005] Step S1: Obtain the power consumption data of each user in the power system at a specified time granularity within the day, form the total load curve of the system, obtain the number of thermal power units, the technical parameters and economic parameters of each thermal power unit, and obtain the predicted output data of each wind farm and photovoltaic power station at a specified time granularity within the day.

[0006] Step S2: Subtract the predicted output values ​​of each wind farm and each photovoltaic power station in each time period from the total load curve of the power system at the specified time granularity within the day to obtain the net load curve of the power system at the specified time granularity within the day. Then, perform the economic dispatch within the day with safety constraints to obtain the dispatch plan of thermal power units and the maximum callable capacity of each thermal power unit in each time period.

[0007] Step S3: Construct the probability distribution of available capacity of each thermal power unit in the two states of "available-out" in each time period, and use discrete recursive convolution to construct the probability distribution of total available thermal power capacity in each time period of the power system. Based on this distribution, perform probability weighted summation on the power deficit when the total available thermal power capacity of all systems in each time period is less than the net load state, calculate the expected value of power shortage EENS in each time period of the system, and accumulate the expected value of total power shortage of the whole system on a time-by-time basis.

[0008] Step S4: Based on the VCG principle, remove each user's load curve from the total system load curve one by one. Repeat steps S2 and S3 to calculate the expected power shortage value of the power system in each time period under the user's absence scenario. Obtain the expected power shortage value of the system in the time series under the user's absence scenario. Subtract the expected power shortage value of the entire system obtained in step S3 from the expected power shortage value of the system in the time series when the user is absent for each time period to quantify the marginal impact of the user's presence or absence on the expected power shortage value of the system in each time period. This will give the user's time series marginal reliability responsibility in each time period. Finally, sum them up to obtain the user's total daily marginal reliability responsibility.

[0009] Step S5: Optimal solution for responsibility allocation. Construct a linear programming model based on the minimum kernel, with the goal of minimizing the stability violation of the maximum proper subset alliance, and solve for the optimal solution for allocating the final reliability responsibility of each user within the day.

[0010] Step S6: Using the proportion of each user's time-series marginal reliability responsibility to their total daily marginal reliability responsibility obtained in Step S4 as the user's time-series weighting factor, the user's intraday final reliability responsibility is allocated to each time period of the day according to its time-series weighting factor, thus obtaining the user's time-series final reliability responsibility in each time period.

[0011] In a preferred embodiment, step S2 specifically includes:

[0012] If we assume that the power consumption of each user is constant at a given time granularity, then the total load of the power system in time period t is considered a fixed value. :

[0013]

[0014] in, Let be the power consumption of user d during time period t; N is the total number of users in the power system.

[0015] During time period t, the predicted output values ​​for each wind farm and photovoltaic power station are: , The total load of the power system in time period t Subtracting the total predicted output of S wind farms and G photovoltaic power plants in time period t, we obtain the net load of the power system in time period t before the dispatch of thermal power units. This constitutes the net load curve of the power system at a specified time granularity within a day:

[0016]

[0017] For the M thermal power units in the power system, implement intraday safety-constrained economic dispatch, with the objective function being to minimize the intraday thermal power generation cost of the system:

[0018]

[0019] Where T represents the total number of periods in the research cycle; ; The unit power generation cost of each thermal power unit obtained in step S1; Let J be the output of thermal power unit j during time period t.

[0020] In a preferred embodiment: system operating constraints include:

[0021] System power balance constraints

[0022]

[0023] Unit output upper and lower limit constraints

[0024]

[0025] in, For the minimum technical output of thermal power unit j, To maximize the technical output of thermal power unit j;

[0026] Unit ramp-up constraints:

[0027]

[0028] in, , P represents the upward and downward ramp rates of thermal power unit j within 15 minutes. j,t-1Let J be the output of thermal power unit j during time period t-1;

[0029] Rotational spare constraint

[0030]

[0031] in, , These are the minimum and maximum technical outputs of thermal power unit j, respectively. , These represent the positive and negative spinning reserve requirements of the system during time period t, respectively, where M represents the total number of thermal power units;

[0032] This yields the daily scheduling plan for thermal power units for each time period, and calculates the maximum available capacity of each thermal power unit for each time period:

[0033]

[0034] in, Let j be the installed capacity of thermal power unit j.

[0035] In a preferred embodiment, step S3 specifically includes:

[0036] From step S1, the forced shutdown rate of thermal power unit j is... Then its availability rate is Define random variables The available capacity of thermal power unit j in time period t is based on the maximum callable capacity of the unit in time period t. Construct a probability distribution model of available capacity in its two states of "available-outage":

[0037]

[0038] definition This represents the probability that the total available capacity of the j thermal power units before time period t equals X. Based on the probability distribution properties of the sum of independent random variables, the probability distribution of the current system's total available thermal power capacity is derived from the previous... The probability distribution of the total available capacity of the j-th thermal power unit is obtained by discrete convolution with the probability distribution of the available capacity of the j-th thermal power unit:

[0039]

[0040] The probability distribution model of available capacity of a single thermal power unit j The non-zero value is only when and When the probability is not zero, substituting it into the above convolution formula, the above summation formula simplifies to the following recursive relationship:

[0041]

[0042]

[0043] in, The initial state of the system without any thermal power units is given by the probability that the total available capacity of the system is 0, which is 1; where, When the j-th thermal power unit is in operation, the j-th thermal power unit is compared with the previous one. Joint probability contribution of available capacity of thermal power units; if the current total available capacity target value X of the system is less than the maximum callable capacity of thermal power unit j. Then the former Total available capacity of Taiwan thermal power units The value must be negative to satisfy the target value of X for the total available capacity of the current system, which violates the physical constraint that the available capacity of thermal power units is non-negative. Therefore, the probability of such an impossible event is 0. Thus, if... Then take .

[0044] In a preferred embodiment: after recursive convolution of M thermal power units, the probability distribution of the total available thermal power capacity of the power system during time period t is obtained. , representing the probability that the total available generating capacity of the system in time period t is X, and its domain is... This is the set of combinations of the maximum callable capacity of all thermal power units;

[0045] Based on the above time period t, the probability distribution of the total available thermal power capacity of the power system Combined with the net load of the power system in time period t Perform a system supply and demand balance analysis; when the total available thermal power capacity X of the system is less than the system net load. When the system power supply is insufficient, the power deficit in this state is: Under a specified time granularity, the total system power consumption during time period t is less than the expected value EENS. t This is the weighted sum of the power deficit under all possible power outage conditions during the period and their probability of occurrence, multiplied by the specified duration of the period:

[0046]

[0047] in, The duration is specified for a given period of time;

[0048] The expected power shortage of the power system at different times of the day (EENS) t Summing these values ​​yields the daily total electricity consumption shortfall in the expected value EENS. total :

[0049]

[0050] Where T represents the total number of periods in the research cycle.

[0051] In a preferred embodiment, step S4 specifically includes:

[0052] Let all users in the power system be set as For any user Construct a system total load curve with specified time granularity within the day after removing the user's load. :

[0053]

[0054] Where T represents the total number of time periods in the research period. ; Let d be the power consumption of user d during time period t;

[0055] After removing user d, the total load of the power system during time period t Subtracting the total predicted output of S wind farms and G photovoltaic power plants in time period t, we obtain the net load of the power system in time period t. This constitutes the net load curve of the power system at a specified time granularity within a day; based on the system net load curve after removing user d. Repeat step S2 to perform the economical scheduling of thermal power units under safety constraints, and obtain the thermal power unit scheduling plan and the maximum available capacity of each thermal power unit in each time period under this scenario. Repeat step S3 to perform the recursive convolution of the thermal power unit, and calculate the expected value of insufficient power in the system at each time period under the user d absence scenario. ;

[0056] Define the time-series marginal reliability responsibility of user d in time period t. The expected value of the total system power shortage EENS obtained in step S3 during time period t. t Expected battery level under time period t in the scenario where user d is absent The difference:

[0057]

[0058] To characterize the marginal impact of the user on the system reliability risk during that period;

[0059] The total daily marginal reliability responsibility of user d is obtained by summing the temporal marginal reliability responsibilities of user d for each time period. :

[0060]

[0061] In a preferred embodiment, step S5 specifically includes:

[0062] A grand alliance is defined as a cooperative alliance formed by all users of a power system without their own subjective will. N is the total number of users in the power system, and any set of users is a non-empty proper subset consortium. satisfy Let the characteristic function For user non-empty proper subset alliance When running independently, only consider The total daily power consumption of the system generated by the user load is less than the expected value. ,definition This is the intraday reliability responsibility allocation vector for the entire system, i.e., the large alliance D users. ,in User d is responsible for intraday reliability.

[0063] Consortium rationality requires that for any user, a non-empty proper subset of the consortium The total reliability responsibility it bears under the major league sharing mechanism It should not exceed The system's daily total power consumption is less than expected when operating independently. That is, it should satisfy ;like ,but Taking on additional responsibilities beyond their independent operating costs theoretically provides an inherent motivation to leave the major leagues.

[0064] In a preferred embodiment: introduce a stability violation vector for proper subset alliances. , This characterizes the excesses beyond the rational boundaries undertaken by each user's non-empty true subset alliance. The additional responsibility; by constructing a linear programming model based on the minimum kernel, with the goal of minimizing the maximum value of the stability violation of the proper subset alliance, the aim is to reduce the maximum unfairness suffered by all user non-empty proper subset alliances to the minimum, thereby finding a set of equilibrium solutions with optimal apportionment fairness and the highest stability of the large alliance under the constraint of ensuring overall balance.

[0065] Objective function:

[0066]

[0067] The objective function aims to find the intraday reliability responsibility allocation vector for users of Major League D. The optimal solution makes the stability of the proper subset alliance violate the degree vector. Minimize the maximum value in;

[0068] Constraints:

[0069] 1) Rational constraints of proper subset alliances

[0070] Union of nonempty proper subsets for any user The total reliability responsibility it bears under the major leagues' cost-sharing mechanism should not exceed [a certain percentage]. Independent risk value and the stability violation currently experienced by the alliance. sum:

[0071]

[0072] in, Indicates the current Under the cost-sharing scheme, the burden borne by the alliance exceeds its rational boundaries. Additional liability;

[0073] 2) Overall equilibrium constraints

[0074] The sum of the daily reliability responsibilities of all users in the entire system must equal the expected value of the total daily power consumption of the entire system obtained in step S3 (EENS). total :

[0075]

[0076] By solving the above linear programming model, the optimal solution vector is obtained. ,in This means that user d bears the ultimate responsibility for the overall reliability of the day. .

[0077] In a preferred embodiment, step S6 specifically includes:

[0078] To scientifically allocate the final reliability responsibility for each user within a day, as determined in step S5, to different time periods, and to reflect the degree of reliability risk causation for users during different periods such as peak and off-peak times, a time-series weighted mapping method is used for allocation; the time-series weighting factor for user d in time period t is calculated. This factor is determined by the proportion of the time-series marginal reliability liability obtained in step S4 to its total daily marginal reliability liability:

[0079]

[0080] Furthermore, the final reliability responsibility for the user within d days obtained in step S5 will be determined. Based on the aforementioned time-series weighting factors allocated to each time period, the final reliability responsibility of user d in time period t is calculated. :

[0081]

[0082] This yields the final time-series reliability responsibility allocation result, which reflects the impact of user electricity consumption behavior on system reliability risk over a time scale.

[0083] This invention also provides a user reliability responsibility calculation system that takes into account flexibility constraints and fairness optimization, including a processor, a memory, and a bus. The memory stores machine-readable instructions executed by the processor. When the system is running, the processor and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, the user reliability responsibility calculation method that takes into account flexibility constraints and fairness optimization is as described above.

[0084] Compared with the prior art, the present invention has the following beneficial effects:

[0085] 1. This invention achieves deep coupling between system flexibility constraints and probabilistic risks on the scheduling time scale. By constructing a system time-series reliability assessment framework that considers the flexibility constraints of thermal power units at a 15-minute granularity, it overcomes the shortcomings of traditional long-term reliability assessment methods that ignore real-time constraints such as ramp-up capability and minimum technical output under the assumption of available capacity at full static output of thermal power units. This provides a quantitative benchmark that conforms to the actual physical laws of system operation for the subsequent accurate tracing of reliability responsibility to the user side.

[0086] 2. This invention proposes a user reliability allocation mechanism based on "minimum kernel optimization for total amount + marginal responsibility with time sequence". It abandons traditional methods such as simple proportional correction or the computationally complex Shapley value method, and proposes a linear programming model based on minimum kernel. With the goal of minimizing the maximum value of the stability violation of the proper subset alliance, it determines the total daily responsibility of users and seeks the optimal solution for user reliability responsibility allocation that is fair, has the highest stability of the cooperative game alliance, and satisfies the overall balance. At the same time, it uses time-series marginal responsibility as a weighting factor to map the total amount to each time period, reflecting the time-series impact characteristics of user electricity consumption behavior on system reliability risk, and maintaining the guiding role and incentive compatibility of responsibility signals on users' optimized electricity consumption behavior. Attached Figure Description

[0087] Figure 1 The flowchart of the user reliability responsibility calculation in this embodiment of the invention takes into account flexibility constraints and fairness optimization.

[0088] Figure 2 The power consumption data of each user at a 15-minute time granularity in this embodiment of the invention.

[0089] Figure 3 Power system time-series supply and demand balance diagram according to an embodiment of the present invention.

[0090] Figure 4 The predicted output values ​​of each wind farm at a 15-minute time granularity in this embodiment of the invention.

[0091] Figure 5 The predicted output values ​​of each photovoltaic power station at a 15-minute time granularity in this embodiment of the invention.

[0092] Figure 6 Thermal diagrams of the dispatched output and sequential output of each thermal power unit in this embodiment of the invention.

[0093] Figure 7 The EENS values ​​of the power system at different time periods under a 15-minute time granularity in this embodiment of the invention.

[0094] Figure 8 The timing reliability responsibility curves for each user in this embodiment of the invention. Detailed Implementation

[0095] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0096] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0097] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations according to this application; as used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise; furthermore, it should be understood that when the terms “comprising” and / or “including” are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0098] refer to Figure 1-8 A method for calculating user reliability responsibility that takes into account flexibility constraints and fairness optimization includes the following steps:

[0099] Step S1: Obtain the daily power consumption data of each user in the power system at a 15-minute time granularity, form the total load curve of the system, obtain the number of thermal power units, the technical parameters (installed capacity, forced outage rate, maximum / minimum technical output, ramp rate, etc.) and economic parameters (unit power generation cost, etc.) of each thermal power unit, and obtain the predicted output data of each wind farm and photovoltaic power station at a 15-minute time granularity within the day;

[0100] Step S2: Subtract the predicted output values ​​of each wind farm and each photovoltaic power station in each time period from the total load curve of the power system at the 15-minute time granularity of the day to obtain the net load curve of the power system at the 15-minute time granularity of the day. Then, perform the daily safety-constrained economic dispatch to obtain the thermal power unit dispatch plan and the maximum callable capacity of each thermal power unit in each time period.

[0101] Step S3: Construct the probability distribution of available capacity of each thermal power unit in the two states of "available-out" in each time period, and use discrete recursive convolution to construct the probability distribution of total available thermal power capacity in each time period of the power system. Based on this distribution, perform probability-weighted summation on the power deficit when the total available thermal power capacity of all systems in each time period is less than the net load state, calculate the expected energy not served (EENS) value of the system in each time period, and accumulate the expected energy not served by the entire system for each time period to obtain the daily total expected energy not served value of the entire system.

[0102] Step S4: Based on the VCG (Vickrey-Clarke-Groves) principle, remove each user's load curve from the total system load curve one by one. Repeat steps S2 and S3 to calculate the expected power shortage value of the power system in each time period under the user's absence scenario. Obtain the expected power shortage value of the system in the time series under the user's absence scenario. Subtract the expected power shortage value of the entire system obtained in step S3 from the expected power shortage value of the system in the time series when the user is absent for each time period to quantify the marginal impact of the user's presence or absence on the expected power shortage value of the system in each time period. This will give the user's time series marginal reliability responsibility in each time period. Finally, sum them up to obtain the user's total daily marginal reliability responsibility.

[0103] Step S5: To find the optimal solution for reliability responsibility allocation that is fair, has the highest stability of the cooperative game alliance, and satisfies overall balance, a linear programming model based on the minimum kernel is constructed. With the goal of minimizing the stability violation of the maximum proper subset alliance, the optimal solution for the allocation of the final reliability responsibility of each user within a day is solved.

[0104] Step S6: Using the proportion of each user's time-series marginal reliability responsibility to their total daily marginal reliability responsibility obtained in Step S4 as the user's time-series weighting factor, the user's intraday final reliability responsibility is allocated to each time period of the day according to its time-series weighting factor, thus obtaining the user's time-series final reliability responsibility in each time period.

[0105] The specific content of step S2 is as follows:

[0106] If we assume that the power consumption of each user is constant at a 15-minute time granularity, then the total load of the power system in time period t is considered a fixed value. :

[0107]

[0108] in, Let d be the power consumption of user d during time period t; N is the total number of users in the power system.

[0109] During time period t, the predicted output values ​​for each wind farm and photovoltaic power station are: , The total load of the power system in time period t Subtracting the total predicted output of S wind farms and G photovoltaic power plants in time period t, we obtain the net load of the power system in time period t before the dispatch of thermal power units. This constitutes the net load curve of the power system at a 15-minute time granularity within a day:

[0110]

[0111] For the M thermal power units in the power system, implement intraday safety-constrained economic dispatch, with the objective function being to minimize the intraday thermal power generation cost of the system:

[0112]

[0113] Where T represents the total number of periods in the research cycle; ; The unit power generation cost of each thermal power unit obtained in step S1; Let J be the output of thermal power unit j during time period t.

[0114] System operational constraints include:

[0115] System power balance constraints

[0116]

[0117] Unit output upper and lower limit constraints

[0118]

[0119] in, For the minimum technical output of thermal power unit j, To maximize the technical output of thermal power unit j.

[0120] Unit ramp-up constraints:

[0121]

[0122] in, , The upward and downward ramp rates (MW / 15min) of thermal power unit j within 15 minutes are respectively.

[0123] Rotational spare constraint

[0124]

[0125] in, , These are the minimum and maximum technical outputs of thermal power unit j, respectively. , These represent the positive and negative rotating reserve requirements of the system during time period t, respectively, and M represents the total number of thermal power units.

[0126] This yields the daily scheduling plan for thermal power units for each time period, and calculates the maximum available capacity of each thermal power unit for each time period:

[0127]

[0128] in, Let j be the installed capacity of thermal power unit j.

[0129] The specific content of step S3 is as follows:

[0130] From step S1, the forced shutdown rate of thermal power unit j is... Then its availability rate is Define random variables. The available capacity of thermal power unit j in time period t is based on the maximum callable capacity of the unit in time period t. Construct a probability distribution model of available capacity in its two states of "available-outage":

[0131]

[0132] definition This represents the probability that the total available capacity of the j thermal power units before time period t equals X. Based on the probability distribution properties of the sum of independent random variables, the probability distribution of the current system's total available thermal power capacity is derived from the previous... The probability distribution of the total available capacity of the j-th thermal power unit is obtained by discrete convolution with the probability distribution of the available capacity of the j-th thermal power unit:

[0133]

[0134] The probability distribution model of available capacity of a single thermal power unit j Non-zero values ​​(i.e., only when...) and Substituting the probability (when not zero) into the above convolution formula, the summation formula simplifies to the following recursive relationship:

[0135]

[0136]

[0137] in, The initial state of the system without any thermal power units is defined as follows: the probability that the total available capacity of the system is 0 is 1. Where, When the j-th thermal power unit is in operation, the j-th thermal power unit is compared with the previous one. The joint probability contribution of the available capacity of the thermal power units. If the current total available capacity target value X of the system is less than the maximum callable capacity of thermal power unit j. Then the former Total available capacity of Taiwan thermal power units The value must be negative to satisfy the target value of X for the total available capacity of the current system, which violates the physical constraint that the available capacity of thermal power units is non-negative. Therefore, the probability of such an impossible event is 0. Thus, if Then take .

[0138] After recursive convolution of M thermal power units, the probability distribution of the total available thermal power capacity of the power system during time period t is obtained. , representing the probability that the total available generating capacity of the system in time period t is X, and its domain is... This is the set of combinations of the maximum callable capacity of all thermal power units.

[0139] Based on the above time period t, the probability distribution of the total available thermal power capacity of the power system Combined with the net load of the power system in time period t Perform a system supply and demand balance analysis. When the total available thermal power capacity X in the system is less than the system net load... When the system power supply is insufficient, the power deficit in this state is: At a 15-minute time granularity, the total system's expected power deficit during time period t is EENS. t This is the weighted sum of the power deficit under all possible power outage conditions during the period and their probability of occurrence, multiplied by the specified duration of the period, which is 15 minutes, or 0.25 hours.

[0140]

[0141] in, The duration is specified for a given period of time;

[0142] The expected power shortage of the power system at different times of the day (EENS) t Summing these values ​​yields the daily total electricity consumption shortfall in the expected value EENS. total :

[0143]

[0144] Where T represents the total number of periods in the research cycle.

[0145] The specific content of step S4 is as follows:

[0146] Given that the VCG mechanism has advantages over cooperative game-theoretic risk-sharing strategies such as the Shapley value method, including lower computational complexity and stronger incentive compatibility, a user absence scenario is constructed based on the VCG principle to quantify the physical marginal impact of each user's electricity consumption behavior on the system's time-series reliability risk. Let the set of all users in the power system be... For any user Construct a total load curve of the system at a 15-minute time granularity within the day after removing the user's load. :

[0147]

[0148] Where T represents the total number of time periods in the research period. ; Let d be the power consumption of user d during time period t.

[0149] After removing user d, the total load of the power system during time period t Subtracting the total predicted output of S wind farms and G photovoltaic power plants in time period t, we obtain the net load of the power system in time period t. This constitutes the net load curve of the power system at a 15-minute time granularity within a day. The system net load curve is based on the removal of user d. Repeat step S2 to perform the economical scheduling of thermal power units under safety constraints, and obtain the thermal power unit scheduling plan and the maximum available capacity of each thermal power unit in each time period under this scenario. Repeat step S3, which involves recursive convolution of the thermal power unit, to calculate the expected power shortage value of the system at each time period under the scenario of user d's absence. .

[0150] Define the time-series marginal reliability responsibility of user d in time period t. The expected value of the total system power shortage EENS obtained in step S3 during time period t. t Expected battery level under time period t in the scenario where user d is absent The difference:

[0151]

[0152] This is to characterize the marginal impact of the user on the system reliability risk during that period.

[0153] The total daily marginal reliability responsibility of user d is obtained by summing the temporal marginal reliability responsibilities of user d for each time period. :

[0154]

[0155] The specific content of step S5 is as follows:

[0156] In actual power system operation, all users form a cooperative alliance without their own subjective will. To achieve a fair distribution of responsibility for the overall system reliability among all users, cooperative game theory is introduced to improve the fairness of reliability responsibility sharing and reduce the overlap of reliability responsibilities among users. The set of all users in the power system is defined as the "large alliance." N is the total number of users in the power system, and any set of users is a non-empty proper subset consortium. satisfy Let the characteristic function For user non-empty proper subset alliance When running independently (i.e., only considering) When the user load is within the system, the total daily power consumption generated is less than the expected value. ,definition This is the intraday reliability responsibility allocation vector for the entire system, i.e., the large alliance D users. ,in Responsibility for user d's intraday reliability.

[0157] In reliability responsibility sharing games, the more reliability responsibility a user assumes, the more reliability costs they must bear. Therefore, federation rationality requires that for any non-empty proper subset of users, the federation... The total reliability responsibility it bears under the major league sharing mechanism It should not exceed The system's daily total power consumption is less than expected when operating independently. That is, it should satisfy .like ,but Having assumed additional responsibilities beyond their independent operating costs, they theoretically have an inherent motivation to leave the major leagues, leading to a decline in the stability of the major leagues.

[0158] To find the fairest, most stable, and strictly balanced optimal solution for the daily reliability responsibility allocation of users, this invention introduces a proper subset alliance stability violation vector. , This characterizes the excesses beyond the rational boundaries undertaken by each user's non-empty true subset alliance. The additional responsibility. By constructing a linear programming model based on the minimum kernel, with the goal of minimizing the maximum value of the stability violation of the proper subset alliance, the aim is to reduce the maximum unfairness suffered by all user non-empty proper subset alliances to the minimum, thereby finding a set of equilibrium solutions with optimal sharing fairness and the highest stability of the large alliance under the constraint of ensuring overall balance.

[0159] (1) Objective function:

[0160]

[0161] The objective function aims to find the intraday reliability responsibility allocation vector for users of Major League D. The optimal solution makes the stability of the proper subset alliance violate the degree vector. The goal is to minimize the maximum value in the large league system, thereby maximizing the relative fairness and stability of the league.

[0162] (2) Constraints:

[0163] 1) Rational constraints of proper subset alliances

[0164] Union of nonempty proper subsets for any user The total reliability responsibility it bears under the major leagues' cost-sharing mechanism should not exceed [a certain percentage]. Independent risk value and the stability violation currently experienced by the alliance. sum:

[0165]

[0166] in, Indicates the current Under the cost-sharing scheme, the burden borne by the alliance exceeds its rational boundaries. The additional workload.

[0167] 2) Overall equilibrium constraints

[0168] The sum of the daily reliability responsibilities of all users in the entire system must equal the expected value of the total daily power consumption of the entire system obtained in step S3 (EENS). total :

[0169]

[0170] By solving the above linear programming model, the optimal solution vector is obtained. ,in This means that user d bears the ultimate responsibility for the overall reliability of the day. .

[0171] The specific content of step S6 is as follows:

[0172] To scientifically allocate the final daily reliability responsibility for each user determined in step S5 to different time periods, and to reflect the degree of reliability risk causation for users during different periods such as peak and off-peak times, a time-series weighted mapping method is used for allocation. The time-series weighting factor for user d in time period t is calculated. This factor is determined by the proportion of the time-series marginal reliability liability obtained in step S4 to its total daily marginal reliability liability:

[0173]

[0174] Furthermore, the final reliability responsibility for the user within d days obtained in step S5 will be determined. Based on the aforementioned time-series weighting factors allocated to each time period, the final reliability responsibility of user d in time period t is calculated. :

[0175]

[0176] This yields the final time-series reliability responsibility allocation result, which reflects the impact of user electricity consumption behavior on system reliability risk over a time scale.

[0177] A preferred embodiment is illustrated in the following specific example:

[0178] A 24-hour period was selected as the research period, and a 15-minute time granularity was used, dividing the data into 96 time periods. On the user side, 15-minute granular power consumption data representing six typical user types were provided, such as... Figure 2 As shown, the total load of the power system for each time period is superimposed, such as... Figure 3 As shown. By Figure 3 It can be seen that the morning peak load periods are 9:00 and 10:00, and the evening peak load periods are 19:00 and 20:00. The power supply side of the power system includes 5 thermal power units, and the parameters of each thermal power unit are shown in Table 1.

[0179] Table 1 Parameters of each thermal power unit

[0180]

[0181] The power system comprises four wind farms and two photovoltaic power stations. The output forecast values ​​for each wind farm and photovoltaic power station are obtained at a 15-minute time granularity. , like Figure 4 , Figure 5 As shown. Based on the power consumption of users and the output forecast data of each wind farm and photovoltaic power station during different time periods of the day, the net load curve of the system at a 15-minute time granularity during the day is obtained as follows. Figure 3 As shown, this is used to execute the intraday safety-constrained economic dispatch of thermal power units, where the spinning reserve demand for each time period is taken as 10% of the net load of the system for each time period, thus obtaining the thermal power unit dispatch plan and the maximum callable capacity of each thermal power unit for each time period, as follows. Figure 6 As shown.

[0182] Based on the maximum available capacity of each thermal power unit in each time period, discrete recursive convolution is used to construct the probability distribution of the total available thermal power capacity of the power system in each time period, and the expected value of power shortage EENS in each time period of the system is calculated. t The system-wide timing EENS curve is formed as follows: Figure 7 As shown, EENS t The peak occurred during the evening rush hour at 75, amounting to 0.0453 MWh. Summing this up, the total daily electricity consumption was less than the expected value EENS.total It is 0.3353 MWh.

[0183] Furthermore, based on the VCG principle, the expected power shortage value of the power system in each time period under the scenario of user absence is calculated, and the system time-series EENS curve under the scenario of user absence is obtained. The difference between the above-obtained system-wide time-series expected power shortage value and the system time-series expected power shortage value when each user is absent is calculated for each time period to obtain the time-series marginal reliability responsibility of each user in each time period. Figure 8 As shown in Table 2, the total daily marginal reliability responsibility of the user is obtained by summing the values, which is 1.1061 MWh (EENS). total Three times that of the previous day. Using a modified linear programming model based on minimum kernel, the final daily reliability responsibility allocation results for each user were obtained, as shown in Table 2.

[0184] Table 2 User Reliability Responsibility Allocation Results

[0185]

[0186] The final allocation results show that the linear programming correction model based on minimum kernel proposed in this invention not only effectively achieves the user reliability responsibility allocation with optimal fairness, the highest stability of the cooperative game alliance, and overall balance, but also demonstrates advantages in suppressing "free-riding" behavior and eliminating unreasonable cross-subsidies. The daily total marginal reliability responsibility of low-peak user 1 and low-power user 5 with a power consumption of only 1 MW has been significantly reduced after the correction, effectively correcting the phenomenon of these users subsidizing other peak users.

[0187] Using the proportion of each user's time-series marginal reliability responsibility to their total daily marginal reliability responsibility as the time-series weighting factor for that user, the user's intraday final reliability responsibility is allocated to each time period within the day according to their time-series weighting factor, resulting in the user's time-series final reliability responsibility for each time period as follows: Figure 8 As shown.

[0188] Preferably, this embodiment confirms that the invention method successfully realizes the mechanism transformation from "allocation based on electricity consumption" to "allocation based on the cause of reliability risk in a time sequence", accurately identifies the source of risk during peak hours, effectively prevents peak users from "free-riding" by not fully assuming their own reliability responsibilities, and thus establishes a user reliability responsibility allocation mechanism that effectively satisfies overall balance, optimal fairness, the highest alliance stability, and incentive compatibility.

[0189] The above description is only a preferred embodiment of the present invention. All equivalent changes and modifications made within the scope of the claims of the present invention should be included in the scope of the present invention.

Claims

1. A method for calculating user reliability responsibility that takes into account flexibility constraints and fairness optimization, characterized in that: Includes the following steps: Step S1: Obtain the power consumption data of each user in the power system at a specified time granularity within the day, form the total load curve of the system, obtain the number of thermal power units, the technical parameters and economic parameters of each thermal power unit, and obtain the predicted output data of each wind farm and photovoltaic power station at a specified time granularity within the day. Step S2: Subtract the predicted output values ​​of each wind farm and each photovoltaic power station in each time period from the total load curve of the power system at the specified time granularity within the day to obtain the net load curve of the power system at the specified time granularity within the day. Then, perform the economic dispatch within the day with safety constraints to obtain the dispatch plan of thermal power units and the maximum callable capacity of each thermal power unit in each time period. Step S3: Construct the probability distribution of available capacity of each thermal power unit in the two states of "available-out" in each time period, and use discrete recursive convolution to construct the probability distribution of total available capacity of thermal power in each time period of the power system. Based on this distribution, perform probability weighted summation on the power deficit when the total available capacity of thermal power in all systems in each time period is less than the net load state, calculate the expected value of power shortage EENS in each time period of the system, and accumulate the expected value of total power shortage of the whole system on a daily basis. Step S4: Based on the VCG principle, remove each user's load curve from the total system load curve one by one. Repeat steps S2 and S3 to calculate the expected power shortage value of the power system in each time period under the user's absence scenario. Obtain the expected power shortage value of the system in the time series under the user's absence scenario. Subtract the expected power shortage value of the entire system obtained in step S3 from the expected power shortage value of the system in the time series when the user is absent for each time period to quantify the marginal impact of the user's presence or absence on the expected power shortage value of the system in each time period. This will give the user's time series marginal reliability responsibility in each time period. Finally, sum them up to obtain the user's total daily marginal reliability responsibility. Step S5: Construct a linear programming model based on the minimum kernel, with the goal of minimizing the stability violation of the maximum proper subset alliance, and solve for the optimal solution for allocating the final reliability responsibility of each user within the day; Step S6: Using the proportion of each user's time-series marginal reliability responsibility to their total daily marginal reliability responsibility obtained in Step S4 as the user's time-series weighting factor, the user's intraday final reliability responsibility is allocated to each time period of the day according to its time-series weighting factor, thus obtaining the user's time-series final reliability responsibility in each time period.

2. The user reliability responsibility calculation method considering flexibility constraints and fairness optimization according to claim 1, characterized in that: The specific content of step S2 is as follows: If we assume that the power consumption of each user is constant at a given time granularity, then the total load of the power system in time period t is considered a fixed value. : in, Let be the power consumption of user d during time period t; N is the total number of users in the power system. During time period t, the predicted output values ​​for each wind farm and photovoltaic power station are: , The total load of the power system in time period t Subtracting the total predicted output of S wind farms and G photovoltaic power plants in time period t, we obtain the net load of the power system in time period t before the dispatch of thermal power units. This constitutes the net load curve of the power system at a specified time granularity within a day: For the M thermal power units in the power system, implement intraday safety-constrained economic dispatch, with the objective function being to minimize the intraday thermal power generation cost of the system: Where T represents the total number of periods in the research cycle; ; The unit power generation cost of each thermal power unit obtained in step S1; Let J be the output of thermal power unit j during time period t.

3. The user reliability responsibility calculation method considering flexibility constraints and fairness optimization according to claim 2, characterized in that: System operational constraints include: System power balance constraints Unit output upper and lower limit constraints in, For the minimum technical output of thermal power unit j, To maximize the technical output of thermal power unit j; Unit ramp-up constraints: in, , P represents the upward and downward ramp rates of thermal power unit j within 15 minutes. j,t-1 Let J be the output of thermal power unit j during time period t-1; Rotational spare constraint in, , These are the minimum and maximum technical outputs of thermal power unit j, respectively. , These represent the positive and negative spinning reserve requirements of the system during time period t, respectively, where M represents the total number of thermal power units; This yields the daily scheduling plan for thermal power units for each time period, and calculates the maximum available capacity of each thermal power unit for each time period: in, Let j be the installed capacity of thermal power unit j.

4. The user reliability responsibility calculation method considering flexibility constraints and fairness optimization according to claim 1, characterized in that: The specific content of step S3 is as follows: From step S1, the forced outage rate of thermal power unit j is... Then its availability rate is Define random variables The available capacity of thermal power unit j in time period t is based on the maximum callable capacity of the unit in time period t. Construct a probability distribution model of available capacity in its two states of "available-out of service": definition Let represent the probability that the total available capacity of j thermal power units before time period t equals X. Based on the probability distribution property of the sum of independent random variables, the probability distribution of the current system's total available thermal power capacity is derived from the previous... The probability distribution of the total available capacity of the j-th thermal power unit is obtained by discretizing and convolving it with the probability distribution of the available capacity of the j-th thermal power unit. The probability distribution model of available capacity of a single thermal power unit j The non-zero value is only when and When the probability is not zero, substituting it into the above convolution formula, the above summation formula simplifies to the following recursive relationship: in, The initial state of the system without any thermal power units is given by the probability that the total available capacity of the system is 0, which is 1; where, When the j-th thermal power unit is in operation, the j-th thermal power unit is compared with the previous one. Joint probability contribution of available capacity of thermal power units; if the current total available capacity target value X of the system is less than the maximum callable capacity of thermal power unit j. Then the former Total available capacity of Taiwan thermal power units The value must be negative to satisfy the target value of X for the total available capacity of the current system, which violates the physical constraint that the available capacity of thermal power units is non-negative. Therefore, the probability of such an impossible event is 0. Thus, if... Then take .

5. The user reliability responsibility calculation method considering flexibility constraints and fairness optimization according to claim 4, characterized in that: After recursive convolution of M thermal power units, the probability distribution of the total available thermal power capacity of the power system during time period t is obtained. , representing the probability that the total available generating capacity of the system in time period t is X, and its domain is... This is the set of combinations of the maximum callable capacity of all thermal power units; Based on the above time period t, the probability distribution of the total available thermal power capacity of the power system Combined with the net load of the power system in time period t Perform a system supply and demand balance analysis; when the total available thermal power capacity X of the system is less than the system net load. When the system power supply is insufficient, the power deficit in this state is: ; At a specified time granularity, the total system's power consumption during time period t is less than the expected value EENS. t This is the weighted sum of the power deficit under all possible power outage conditions during the period and their probability of occurrence, multiplied by the specified duration of the period: in, The duration is specified for a given period of time; The expected power shortage of the power system at different times of the day (EENS) t Summing these values ​​yields the daily total electricity consumption shortfall in the expected value EENS. total : Where T represents the total number of periods in the research cycle.

6. The user reliability responsibility calculation method considering flexibility constraints and fairness optimization according to claim 1, characterized in that: The specific content of step S4 is as follows: Let all users in the power system be set as For any user Construct a system total load curve with specified time granularity within the day after removing the user's load. : Where T represents the total number of time periods in the research period. ; The power consumption of user d during time period t; After removing user d, the total load of the power system during time period t Subtracting the total predicted output of S wind farms and G photovoltaic power plants in time period t, we obtain the net load of the power system in time period t. This constitutes the net load curve of the power system at a specified time granularity within a day; based on the system net load curve after removing user d. Repeat step S2 to perform the economical scheduling of thermal power units under safety constraints, and obtain the thermal power unit scheduling plan and the maximum available capacity of each thermal power unit in each time period under this scenario. Repeat step S3 to perform the recursive convolution of the thermal power unit, and calculate the expected value of insufficient power in the system at each time period under the user d absence scenario. ; Define the time-series marginal reliability responsibility of user d in time period t. The expected value of the total system power shortage EENS obtained in step S3 during time period t. t Expected battery level under time period t in the scenario where user d is absent The difference: To characterize the marginal impact of the user on the system reliability risk during that period; The total daily marginal reliability responsibility of user d is obtained by summing the temporal marginal reliability responsibilities of user d for each time period. : 。 7. The user reliability responsibility calculation method considering flexibility constraints and fairness optimization according to claim 1, characterized in that: The specific content of step S5 is as follows: A grand alliance is defined as a cooperative alliance formed by all users of a power system without their own subjective will. N is the total number of users in the power system, and any set of users is a non-empty proper subset consortium. satisfy ; Let the characteristic function For user non-empty proper subset alliance When running independently, only consider The total daily power consumption of the system generated by the user load is less than the expected value. ,definition This is the intraday reliability responsibility allocation vector for the entire system, i.e., the large alliance D users. ,in User d is responsible for intraday reliability. Consortium rationality requires that for any user, a non-empty proper subset of the consortium The total reliability responsibility it bears under the major league sharing mechanism It should not exceed The system's daily total power consumption is less than expected when operating independently. That is, it should satisfy ;like ,but Taking on additional responsibilities beyond their independent operating costs theoretically provides an inherent motivation to leave the major leagues.

8. The user reliability responsibility calculation method considering flexibility constraints and fairness optimization according to claim 1, characterized in that: Introducing a stability violation degree vector for proper subset alliances , This characterizes the excesses beyond the rational boundaries undertaken by each user's non-empty true subset alliance. The additional responsibility; by constructing a linear programming model based on the minimum kernel, with the goal of minimizing the maximum value of the stability violation of the proper subset alliance, the aim is to reduce the maximum unfairness suffered by all user non-empty proper subset alliances to the minimum, thereby finding a set of equilibrium solutions with optimal apportionment fairness and the highest stability of the large alliance under the constraint of ensuring overall balance. Objective function: The objective function aims to find the intraday reliability responsibility allocation vector for users of Major League D. The optimal solution makes the stability of the proper subset alliance violate the degree vector. Minimize the maximum value in; Constraints: 1) Rational constraints of proper subset alliances Union of nonempty proper subsets for any user The total reliability responsibility it bears under the major leagues' cost-sharing mechanism should not exceed [a certain percentage]. Independent risk value and the stability violation currently experienced by the alliance. sum: in, Indicates the current Under the cost-sharing scheme, the burden borne by the alliance exceeds its rational boundaries. Additional liability; 2) Overall equilibrium constraints The sum of the daily reliability responsibilities of all users in the entire system must equal the expected value of the total daily power consumption of the entire system obtained in step S3 (EENS). total : By solving the above linear programming model, the optimal solution vector is obtained. ,in This means that user d bears the ultimate responsibility for the overall reliability of the day. .

9. The user reliability responsibility calculation method considering flexibility constraints and fairness optimization according to claim 1, characterized in that: The specific content of step S6 is as follows: To scientifically allocate the final reliability responsibility of each user within a day as determined in step S5 to each time period, so as to reflect the degree of reliability risk causation of users in different time periods such as peak and off-peak periods, a time-series weighted mapping method is used for allocation. Calculate the time-series weighting factor for user d in time period t. This factor is determined by the proportion of the time-series marginal reliability liability obtained in step S4 to its total daily marginal reliability liability: Furthermore, the final reliability responsibility for the user within d days obtained in step S5 will be determined. Based on the aforementioned time-series weighting factors allocated to each time period, the final reliability responsibility of user d in time period t is calculated. : This yields the final time-series reliability responsibility allocation result, which reflects the impact of user electricity consumption behavior on system reliability risk over a time scale.

10. A user reliability liability calculation system considering flexibility constraints and fairness optimization, comprising a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executed by the processor; characterized in that, When the system is running, the processor and the memory communicate via a bus, and the machine-readable instructions are executed by the processor as described in any one of claims 1 to 9, which is a user reliability responsibility calculation method that takes into account flexibility constraints and fairness optimization.

Images