Power transaction method and system based on multi-party security calculation and federated learning

Abstract

The invention discloses a power transaction method and system based on multi-party security calculation and federated learning, and the method comprises the steps: obtaining a power transaction region, and obtaining the historical monitoring data of a power utilization unit corresponding to each regional power utilization unit in the power transaction region; dividing regional power utilization units in the power transaction region into first type power utilization units and second type power utilization units according to the historical monitoring data of the power utilization units; according to the power consumption unit type division result, analyzing and obtaining time period estimated power consumption, a price sensitivity evaluation coefficient and a power supply index value, and obtaining time period power consumption analysis data; dividing the plurality of power consumption characteristic time periods into a first type of power consumption time periods and a second type of power consumption time periods according to the time period power consumption analysis data, and performing power transaction control on regional power consumption units according to a power consumption unit type division result and a power consumption characteristic time period division result; compared with a traditional method, the economic benefit of an electric power company and the safety of residential electricity utilization can be improved.

Description

Technical Field

[0001] The present invention relates to an electricity trading method and system, and in particular to an electricity trading method and system based on multi-party secure computing and federated learning, belonging to the technical field of electricity trading. Background Art

[0002] The existing electricity trading methods have the following specific defects when conducting electricity trading: the existing electricity trading methods often adopt a fixed peak-valley electricity price trading method. This electricity trading method cannot respond to changes in market supply and demand in a timely manner, and cannot make real-time electricity trading adjustments based on the electricity consumption of electricity users in different time periods, which leads to poor electricity regulation and control effects and a certain amount of waste of electricity resources; the existing electricity trading methods are difficult to achieve effective monitoring and management of users' electricity consumption behaviors, and cannot avoid malicious electricity consumption by adjusting the electricity trading methods, which leads to hidden dangers in the operating interests of power companies and the safety of residents' electricity use. Summary of the invention

[0003] Purpose of the invention: The purpose of the present invention is to provide an electricity trading method and system based on multi-party secure computing and federated learning that can reduce the waste of electricity resources.

[0004] Technical solution: The power trading method based on multi-party secure computing and federated learning described in the present invention includes:

[0005] S1: Obtain the power trading area, and respectively obtain the power unit historical monitoring data corresponding to each power unit in the power trading area; the power unit historical monitoring data includes: abnormal power cycle number ratio, cycle average power consumption deviation and cycle power consumption deviation variance;

[0006] S2: Divide the regional power users in the power trading area into first-type power users and second-type power users according to the historical monitoring data of the power users;

[0007] S3: According to the classification result of the power consumption unit type, a power consumption characteristic period is marked in the future power consumption period of the first type of power consumption unit, and the power consumption characteristic period is analyzed to obtain the estimated power consumption, price sensitivity evaluation coefficient and power supply index value of each first type of power consumption unit in the power consumption characteristic period, so as to obtain the power consumption analysis data of the period;

[0008] S4: Divide a plurality of power consumption characteristic time periods into a first type of power consumption time period and a second type of power consumption time period according to the power consumption analysis data during the time period, and perform power trading control on regional power consumption units according to the power consumption unit type division result and the power consumption characteristic time period division result.

[0009] Furthermore, the step S1 comprises:

[0010] S11: Obtain the power trading area covered by the power company and each regional power user in the power trading area, and randomly select one power user from each regional power user as a sample power user;

[0011] S12: monitoring the power transaction of the sample power users to obtain historical monitoring data of the sample power users;

[0012] S13: Based on the historical monitoring data of the sample electricity consumption units, respectively obtain the historical monitoring data corresponding to each regional electricity consumption unit;

[0013] S14: defining the historical monitoring data corresponding to the power consumption unit in each area as the power consumption unit historical monitoring data.

[0014] Furthermore, the step S12 includes:

[0015] S121: Obtain historical electricity consumption data corresponding to a sample electricity consumption unit, and divide the historical electricity consumption data into a number of historical electricity consumption monitoring periods based on a preset monitoring time period; the historical electricity consumption data includes electricity consumption and electricity power;

[0016] S122: Assume that each historical power consumption monitoring period is from the first historical power consumption monitoring period to the i-th historical power consumption monitoring period;

[0017] S123: Perform unit power analysis on all historical power monitoring cycles of the sample power-consuming units to obtain the ratio of the number of abnormal power cycles; specifically, the following includes:

[0018] In the first historical power consumption monitoring cycle, a number of power monitoring time points are randomly selected, and the obtained power monitoring time points are named as the first power monitoring time point to the jth power monitoring time point respectively;

[0019] Respectively obtain the power consumption values ​​corresponding to the sample power consumption units from the first power monitoring time point to the jth power monitoring time point, and obtain the first power consumption value to the jth power consumption value;

[0020] Obtain the reference power values ​​corresponding to the first power monitoring time point to the jth power monitoring time point in the power trading area, and obtain the first reference power value to the jth reference power value;

[0021] The periodic power deviation corresponding to the first historical power monitoring period is obtained by calculating the first power consumption value to the jth power consumption value and the first reference power consumption value to the jth reference power consumption value. The specific formula is as follows:

[0022]

[0023] Wherein, Wpc is the periodic power deviation corresponding to the first historical power monitoring period, Wsl1 to Wslj are respectively the first power value to the jth power value, Wsj1 to Wsjj are respectively the first reference power value to the jth reference power value, and j is the number of power monitoring time points;

[0024] Respectively obtain the periodic power deviations corresponding to the second historical power monitoring period to the i-th historical power monitoring period to obtain the corresponding periodic power deviations;

[0025] Set a periodic power consumption deviation threshold, compare the power consumption deviation of each period with multiple periodic power consumption deviation thresholds, and when a periodic power consumption deviation is greater than or equal to the periodic power consumption deviation threshold, mark the corresponding historical power consumption monitoring period as an abnormal power monitoring period; when a periodic power consumption deviation is less than the periodic power consumption deviation threshold, mark the corresponding historical power consumption monitoring period as a normal power monitoring period;

[0026] Counting the number of abnormal power monitoring cycles from the first historical power consumption monitoring cycle to the i-th historical power consumption monitoring cycle to obtain a number value of abnormal power monitoring cycles, and calculating a ratio of the number value of abnormal power monitoring cycles to i to obtain a number ratio of abnormal power cycles;

[0027] S124: Obtaining the power consumption of the sample power consumption unit from the first historical period to the i-th historical period;

[0028] S125: Obtain the average power consumption per unit period of all regions in the power trading area from the first historical power consumption monitoring period to the i-th historical power consumption monitoring period, and obtain the benchmark power consumption from the first historical period to the i-th historical period;

[0029] S126: Calculate the difference between the power consumption of each historical period of the sample power consumption unit and the corresponding historical period benchmark power consumption, and obtain the power consumption deviation from the first historical period to the i-th historical period;

[0030] S127: Calculate the average of the power consumption deviation from the first historical period to the power consumption deviation from the i-th historical period to obtain the period average power consumption deviation;

[0031] S128: Calculate the variance of the first historical period power consumption deviation to the i-th historical period power consumption deviation to obtain the period power consumption deviation variance;

[0032] S129: The abnormal power cycle number ratio, cycle average power consumption deviation, and cycle power consumption deviation variance corresponding to the sample power consumption units are defined as historical monitoring data.

[0033] Furthermore, the step S2 comprises:

[0034] S21: Obtaining historical monitoring data corresponding to each regional electricity user;

[0035] S22: Analyze the power consumption data according to the historical monitoring data of the sample power consumption units to obtain the power data abnormality analysis coefficients of the sample power consumption units, and analyze the power consumption data to obtain the power data abnormality analysis coefficients corresponding to the power consumption units in each region;

[0036] S23: Based on the calculation method of the power data anomaly analysis coefficient of the sample power units, the power data anomaly analysis coefficient corresponding to each regional power unit is calculated respectively to obtain the corresponding power data anomaly analysis coefficient; a power data anomaly analysis coefficient threshold corresponding to each regional power unit is set, and each power data anomaly analysis coefficient is numerically compared with the power data anomaly analysis coefficient threshold, and the regional power units are divided into first type power units and second type power units according to the numerical comparison results to obtain power unit type classification data.

[0037] Furthermore, the step S22 includes:

[0038] S221: Obtaining the abnormal power cycle number ratio, cycle average power consumption deviation, and cycle power consumption deviation variance of the sample power consumption unit according to the historical monitoring data corresponding to the sample power consumption unit;

[0039] S222: Calculate the power data abnormality analysis coefficient based on the abnormal power cycle number ratio, cycle average power consumption deviation, and cycle power consumption deviation variance of the sample power consumption unit:

[0040] Dyx=Zyb+Zpp*(1+Zfc)+Zpp

[0041] Among them, Dyx is the power data abnormality analysis coefficient, Zyb is the abnormal power cycle number ratio, Zpp is the cycle average power consumption deviation, and Zfc is the cycle power consumption deviation variance.

[0042] Further, the step S23 includes:

[0043] Step S231: setting the abnormal power cycle number ratio threshold, cycle average power consumption deviation threshold and cycle power consumption deviation variance threshold of each regional power consumption unit respectively;

[0044] Step S232: Calculate the power data abnormality analysis coefficient threshold by using the abnormal power cycle number ratio threshold, the cycle average power consumption deviation threshold, and the cycle power consumption deviation variance threshold;

[0045] Step S233: when the power data abnormality analysis coefficient is greater than or equal to the power data abnormality analysis coefficient threshold, the corresponding regional power consumption unit is classified as a first type of power consumption unit;

[0046] Step S234: when the power data abnormality analysis coefficient is less than the power data abnormality analysis coefficient threshold, the corresponding regional power consumption unit is classified as a second type of power consumption unit.

[0047] Furthermore, the step S3 comprises:

[0048] S31: According to the classification result of the electricity consumption unit type, randomly select one electricity consumption unit from all the first type electricity consumption units as a characteristic electricity consumption unit;

[0049] S32: Preset a time length as a characteristic power usage duration, mark the time point corresponding to the current moment as a first power usage time point, mark a time point corresponding to a characteristic power usage duration after the first power usage time point as a second power usage time point, and mark the period between the first power usage time point and the second power usage time point as a power usage characteristic period;

[0050] S33: Performing time period power analysis on the characteristic time period of the characteristic power consumption unit to obtain time period analysis data corresponding to the characteristic power consumption unit;

[0051] S34: based on the time period analysis data corresponding to the characteristic electricity consumption unit, respectively acquiring the time period analysis data of each first type electricity consumption unit in the characteristic electricity consumption unit, to obtain a plurality of time period analysis data;

[0052] S35: defining the power consumption characteristic time period and the time period power consumption analysis data corresponding to the plurality of first type power consumption units as the time period power consumption analysis data.

[0053] Furthermore, the step S33 includes:

[0054] S331: Marking a power consumption evaluation period in the historical power consumption data corresponding to the characteristic power consumption unit, and obtaining multiple time periods in the power consumption evaluation period that are the same as the power consumption characteristic time period as the power consumption evaluation period, to obtain multiple power consumption evaluation periods;

[0055] S332: respectively obtaining the power consumption of the characteristic power consumption unit in each power consumption evaluation period, obtaining the power consumption of multiple periods, and averaging the obtained power consumption of the multiple periods to obtain the estimated power consumption of the characteristic power consumption unit in the period corresponding to the power consumption characteristic period;

[0056] S333: Obtaining the price sensitivity evaluation coefficient corresponding to the characteristic electricity consumption unit; specifically including:

[0057] Mark multiple time periods in the historical electricity consumption data that are the same as the time range of the electricity consumption characteristic period as price sensitivity evaluation periods, and obtain multiple price sensitivity evaluation periods, and the electricity price corresponding to each price sensitivity evaluation period is different;

[0058] Obtain the unit electricity price of the characteristic electricity user in each price sensitivity assessment period, and sort the obtained unit electricity prices in multiple periods in descending order according to the numerical value to obtain a unit electricity price descending sorting queue;

[0059] In the descending order of the unit electricity price queue, several time period unit electricity prices are selected as characteristic unit electricity prices, and the price difference between two consecutive characteristic unit electricity prices is constrained to be the same, and the electricity price value is obtained for each characteristic unit electricity price to obtain multiple characteristic unit electricity price values, and the obtained multiple characteristic unit electricity price values ​​are named from the first characteristic unit electricity price to the pth unit electricity price in order from small to large;

[0060] Obtain the power consumption of the characteristic power consumption unit in the price sensitivity assessment period from the first characteristic unit electricity price to the pth characteristic unit electricity price, and obtain the first characteristic power consumption to the pth characteristic power consumption;

[0061] In the plane rectangular coordinate system, the characteristic electricity consumption is set as the coordinate y-axis, and the characteristic unit electricity price is set as the coordinate x-axis to obtain the price sensitivity analysis coordinate system;

[0062] In the price sensitivity analysis coordinate system, the first characteristic unit electricity price to the pth characteristic unit electricity price are used as the horizontal coordinate, and the first characteristic power consumption to the pth characteristic power consumption are used as the vertical coordinate to obtain the first characteristic coordinate point to the pth characteristic coordinate point;

[0063] Mark the first characteristic coordinate point to the pth characteristic coordinate point in the price sensitivity analysis coordinate system respectively, and connect each adjacent characteristic coordinate point to obtain p-1 characteristic polylines;

[0064] Obtain the slopes of all p-1 characteristic broken lines, obtain p-1 broken line slopes, and average the obtained p-1 broken line slopes to obtain the price sensitivity assessment coefficient corresponding to the characteristic electricity unit;

[0065] S334: Obtain the estimated power demand and estimated power supply of the power company during the power consumption characteristic period, set the power supply index value as a parameter, compare the estimated power demand and the estimated power supply, when the estimated power demand is greater than or equal to the estimated power supply, the power supply index value is 0, when the estimated power demand is less than the estimated power supply, the power supply index value is the absolute value of the difference between the estimated power supply and the estimated power demand;

[0066] S335: The estimated power consumption during a time period, the price sensitivity evaluation coefficient, and the power supply index value are defined as the time period analysis data corresponding to the characteristic power consumption unit.

[0067] Furthermore, the step S4 comprises:

[0068] S41: Calculate the corresponding electricity consumption period type division coefficient according to the estimated electricity consumption, price sensitivity evaluation coefficient and power supply index value of each first type electricity user unit during the period:

[0069]

[0070] Among them, Sdh is the electricity consumption period type classification coefficient, Mgx is the price sensitivity assessment coefficient, Gdz is the power supply index value, and Ydl is the estimated power consumption during the period;

[0071] S42: Obtaining a power consumption period type division coefficient threshold, comparing the values ​​of each power consumption period type division coefficient with the power consumption period type division coefficient threshold, dividing the power consumption feature period into a first type power consumption period and a second type power consumption period according to the numerical comparison result, and obtaining the power consumption feature period division result; including:

[0072] S421: setting a time period estimated power consumption threshold, a price sensitivity assessment coefficient threshold, and a power supply index threshold respectively;

[0073] S422: Based on the calculation method of the electricity consumption period classification system, the estimated electricity consumption threshold, the price sensitivity evaluation coefficient threshold and the power supply index threshold are calculated to obtain the electricity consumption period classification coefficient threshold;

[0074] S423: if the power consumption period type division coefficient is greater than or equal to the power consumption period type division coefficient threshold, the power consumption characteristic period is divided into a first type of power consumption period;

[0075] S424: if the power consumption period type division coefficient is less than the power consumption period type division coefficient threshold, the power consumption characteristic period is divided into a second type of power consumption period;

[0076] S43: Controlling power trading of regional power users according to the results of power user type classification and power consumption characteristic time period classification; specifically including:

[0077] When the regional electricity user is a first type electricity user and the electricity consumption characteristic period is a first type electricity consumption period, the power company stops conducting electricity transactions with the regional electricity user during the electricity consumption characteristic period;

[0078] When the regional electricity user is a first type electricity user and the characteristic electricity consumption period is a second type electricity consumption period, the power company increases the electricity transaction price of the regional electricity user during the characteristic electricity consumption period;

[0079] When the regional electricity user is a second type of electricity user, the power company normally conducts electricity transactions with the regional electricity user.

[0080] Based on the same inventive concept, the present invention also provides an electric power trading system based on multi-party secure computing and federated learning, comprising:

[0081] The data acquisition module is used to acquire the power trading area and respectively acquire the power unit historical monitoring data corresponding to each power unit in the power trading area; the power unit historical monitoring data includes: abnormal power cycle number ratio, cycle average power consumption deviation and cycle power consumption deviation variance;

[0082] A unit analysis module, used to classify regional power users in the power trading area into first type power users and second type power users according to historical monitoring data of power users;

[0083] The time period analysis module is used to mark a power consumption characteristic time period in the future power consumption time period of the first type of power consumption unit according to the power consumption unit type classification result, and obtain the time period power consumption analysis data by performing power analysis on the power consumption characteristic time period to obtain the time period estimated power consumption, price sensitivity evaluation coefficient and power supply index value corresponding to each first type of power consumption unit in the power consumption characteristic time period;

[0084] The power trading module is used to divide multiple power consumption characteristic time periods into first type power consumption time periods and second type power consumption time periods according to the power consumption analysis data of the time periods, and to control power trading of regional power consumption units according to the power consumption unit type division results and the power consumption characteristic time period division results.

[0085] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: 1. The present invention obtains the abnormal power cycle number ratio, the cycle average power consumption deviation and the cycle power consumption deviation variance to calculate the power data abnormal analysis coefficient to classify the regional power users, and adopts different power trading methods for different types of regional power users, which can make real-time power trading adjustments and effectively reduce the waste of power resources; 2. The present invention obtains the estimated power consumption, price sensitivity assessment coefficient and power supply index value corresponding to the predicted future power consumption period to perform time-based control of the power trading method of the regional power users, which can effectively improve the economic benefits of the power company and the safety of residents' electricity use. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] Figure 1is a flow chart of a method according to an embodiment of the present invention;

[0087] Figure 2 is an overall system block diagram of an embodiment of the present invention;

[0088] Figure 3 Schematic diagram of a price sensitivity analysis coordinate system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0089] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application; it is obvious that the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0090] Embodiment 1: As shown in the attached Figure 1 As shown, the power trading method based on multi-party secure computing and federated learning in this embodiment includes:

[0091] S1: Obtain the power trading area, and respectively obtain the power unit historical monitoring data corresponding to each power unit in the power trading area; the power unit historical monitoring data includes: abnormal power cycle number ratio, cycle average power consumption deviation and cycle power consumption deviation variance;

[0092] S2: Divide the regional power users in the power trading area into first-type power users and second-type power users according to the historical monitoring data of the power users;

[0093] S3: According to the classification result of the power consumption unit type, a power consumption characteristic period is marked in the future power consumption period of the first type of power consumption unit, and the power consumption characteristic period is analyzed to obtain the estimated power consumption, price sensitivity evaluation coefficient and power supply index value of each first type of power consumption unit in the power consumption characteristic period, so as to obtain the power consumption analysis data of the period;

[0094] S4: Divide a plurality of power consumption characteristic time periods into a first type of power consumption time period and a second type of power consumption time period according to the power consumption analysis data during the time period, and perform power trading control on regional power consumption units according to the power consumption unit type division result and the power consumption characteristic time period division result.

[0095] Specifically, step S1 includes:

[0096] Step S11: obtaining the power trading area covered by the power company, obtaining each power user in the power trading area respectively, obtaining multiple regional power users, and randomly selecting one power user from the multiple regional power users as a sample power user;

[0097] Step S12: monitoring the power transactions of the sample power users to obtain the historical monitoring data corresponding to the sample power users;

[0098] The step S12 further includes the following specific steps:

[0099] Step S121: acquiring historical power consumption data corresponding to the sample power consumption unit, marking a number of historical power consumption monitoring cycles with a time length as the characteristic power monitoring time length in the historical power consumption data, and the time interval between each two consecutive historical power consumption monitoring cycles is the same;

[0100] Step S122: naming the acquired multiple historical power consumption monitoring cycles as the first historical power consumption monitoring cycle to the i-th historical power consumption monitoring cycle respectively;

[0101] Step S123: performing unit power analysis on the first historical power monitoring period to the i-th historical power monitoring period to obtain the ratio of the number of abnormal power periods;

[0102] The step S123 further includes the following specific steps:

[0103] Step S1231: randomly selecting a number of power monitoring time points within the first historical power consumption monitoring cycle, and naming the obtained power monitoring time points as the first power monitoring time point to the jth power monitoring time point respectively;

[0104] Step S1232: respectively obtaining the power consumption values ​​of the sample power consumption units corresponding to the first power monitoring time point to the jth power monitoring time point, and obtaining the first power consumption value to the jth power consumption value;

[0105] Step S1233: Obtain the reference power values ​​corresponding to the power trading area at the first power monitoring time point to the jth power monitoring time point, and obtain the first reference power value to the jth reference power value;

[0106] Step S1234: Calculate the periodic power deviation corresponding to the first historical power monitoring period by calculating the first power consumption value to the jth power consumption value and the first reference power consumption value to the jth reference power consumption value;

[0107] The periodic power deviation corresponding to the first historical power monitoring period is calculated, and the specific formula is as follows:

[0108]

[0109] Wherein, Wpc is the periodic power deviation corresponding to the first historical power monitoring period, Wsl1 to Wslj are respectively the first power value to the jth power value, Wsj1 to Wsjj are respectively the first reference power value to the jth reference power value, and j is the quantity value corresponding to the power monitoring time point;

[0110] Step S1235: respectively acquiring the periodic power deviations corresponding to the second historical power monitoring period to the i-th historical power monitoring period to obtain a plurality of periodic power deviations;

[0111] Step S1236: Obtain a periodic power consumption deviation threshold value, compare multiple periodic power consumption deviations with multiple periodic power consumption deviation threshold values ​​respectively, when the periodic power consumption deviation is greater than or equal to the periodic power consumption deviation threshold value, mark the corresponding historical power consumption monitoring period as an abnormal power monitoring period, when the periodic power consumption deviation is less than the periodic power consumption deviation threshold value, mark the corresponding historical power consumption monitoring period as a normal power monitoring period;

[0112] Step S1237: Count the number of the first historical power monitoring cycle to the i-th historical power monitoring cycle to obtain the number of abnormal power monitoring cycles, calculate the ratio of the number of abnormal power monitoring cycles to i, and obtain the number ratio of abnormal power cycles;

[0113] Step S124: respectively acquiring the periodic power consumption of the sample power consumption units from the first historical power consumption monitoring period to the i-th historical power consumption monitoring period, and obtaining the power consumption from the first historical period to the i-th historical period;

[0114] Step S125: respectively obtaining the average power consumption per unit period of all regional power consumptions corresponding to the power trading area in the first historical power consumption monitoring period to the i-th historical power consumption monitoring period, and obtaining the benchmark power consumption of the first historical period to the i-th historical period;

[0115] Step S126: Calculate the difference between the power consumption of the first historical period and the benchmark power consumption of the first historical period to obtain the power consumption deviation of the first period, calculate the difference between the power consumption of the second historical period and the benchmark power consumption of the second historical period to obtain the power consumption deviation of the second period, repeat this process, and obtain the power consumption deviation of the third period to the power consumption deviation of the i-th period respectively;

[0116] Step S127: Calculate the average of the power consumption deviation from the first cycle to the power consumption deviation of the i-th cycle to obtain the cycle average power consumption deviation;

[0117] Step S128: Calculate the variance of the first cycle power consumption deviation to the i-th cycle power consumption deviation to obtain the cycle power consumption deviation variance;

[0118] Step S129: defining the abnormal power cycle number ratio, cycle average power consumption deviation and cycle power consumption deviation variance corresponding to the sample power consumption unit as historical monitoring data;

[0119] Step S13: acquiring the historical monitoring data corresponding to each regional power consumption unit respectively to obtain a plurality of historical monitoring data;

[0120] Step S14: defining a plurality of historical monitoring data as historical monitoring data of a power consumption unit;

[0121] Step S2: dividing the regional power users in the power trading area into first-type power users and second-type power users according to the historical monitoring data of the power users, and obtaining power user type division data;

[0122] Step S21: Obtain historical monitoring data of the power consumption unit, and obtain historical monitoring data corresponding to each regional power consumption unit according to the historical monitoring data of the power consumption unit;

[0123] Step S22: Analyze the power consumption data of the sample power consumption unit according to the historical monitoring data corresponding to the sample power consumption unit, and obtain the power data abnormality analysis coefficient corresponding to the sample power consumption unit;

[0124] The step S22 further includes the following specific steps:

[0125] Step S221: Obtain the abnormal power cycle number ratio, cycle average power consumption deviation, and cycle power consumption deviation variance respectively according to the historical monitoring data corresponding to the sample power consumption unit;

[0126] Step S222: Calculate the power data abnormality analysis coefficient by using the abnormal power cycle number ratio, the cycle average power consumption deviation, and the cycle power consumption deviation variance;

[0127] The power data abnormality analysis coefficient is calculated, and the specific formula is as follows:

[0128] Dyx=Zyb+Zpp*(1+Zfc)+Zpp;

[0129] Among them, Dyx is the power data abnormality analysis coefficient, Zyb is the abnormal power cycle number ratio, Zpp is the cycle average power consumption deviation, and Zfc is the cycle power consumption deviation variance;

[0130] Step S23: respectively obtaining the power data abnormality analysis coefficient corresponding to each regional power consumption unit, obtaining a plurality of power data abnormality analysis coefficients, obtaining a power data abnormality analysis coefficient threshold, numerically comparing each power data abnormality analysis coefficient with the power data abnormality analysis coefficient threshold, and dividing the regional power consumption units into first type power consumption units and second type power consumption units according to the numerical comparison results, to obtain power consumption unit type division data;

[0131] The step S23 further includes the following specific steps:

[0132] Step S231: respectively obtaining an abnormal power cycle number ratio threshold, a cycle average power consumption deviation threshold, and a cycle power consumption deviation variance threshold;

[0133] Step S232: Calculate the power data abnormality analysis coefficient threshold by using the abnormal power cycle number ratio threshold, the cycle average power consumption deviation threshold, and the cycle power consumption deviation variance threshold;

[0134] The power data abnormality analysis coefficient threshold is calculated, and the specific formula is as follows:

[0135] Dyxy=Zyby+Zppy*(1+Zfcy)+Zppy;

[0136] Among them, Dyxy is the power data abnormality analysis coefficient threshold, Zyby is the abnormal power cycle number ratio threshold, Zppy is the cycle average power consumption deviation threshold, and Zfcy is the cycle power consumption deviation variance threshold;

[0137] Step S233: when the power data abnormality analysis coefficient is greater than or equal to the power data abnormality analysis coefficient threshold, the corresponding regional power consumption unit is classified as a first type of power consumption unit;

[0138] Step S234: when the power data abnormality analysis coefficient is less than the power data abnormality analysis coefficient threshold, the corresponding regional power consumption unit is classified as a second type of power consumption unit;

[0139] Step S3: Mark a power consumption characteristic period in the future power consumption period of the first type of power consumption unit according to the power consumption unit type classification data, and obtain the time period estimated power consumption, price sensitivity assessment coefficient and power supply index value corresponding to each first type of power consumption unit by performing power analysis on the power consumption characteristic period, so as to obtain the time period power consumption analysis data;

[0140] Step S31: acquiring power unit type classification data, acquiring each first type power unit in the power trading area according to the power unit type classification data, and randomly selecting one power unit from the acquired multiple first type power units as a characteristic power unit;

[0141] Step S32: Mark the time point corresponding to the current moment as the first power consumption time point, mark the time point corresponding to a characteristic power consumption duration after the first power consumption time point as the second power consumption time point, and mark the period between the first power consumption time point and the second power consumption time point as the power consumption characteristic period;

[0142] Step S33: performing time period power analysis on characteristic power consumption units in the characteristic power consumption time period to obtain time period analysis data corresponding to the characteristic power consumption units;

[0143] The step S33 further includes the following specific steps:

[0144] Step S331: obtaining historical electricity consumption data corresponding to the characteristic electricity consumption unit, marking a power consumption evaluation period in the historical electricity consumption data, and obtaining multiple time periods in the power consumption evaluation period that are the same as the time range of the power consumption characteristic period as the power consumption evaluation period, to obtain multiple power consumption evaluation periods;

[0145] Step S332: respectively obtaining the power consumption of the characteristic power consumption unit in each power consumption evaluation period, obtaining the power consumption of multiple periods, and averaging the obtained power consumption of the multiple periods to obtain the estimated power consumption of the characteristic power consumption unit in the power consumption characteristic period;

[0146] Step S333: Obtaining the price sensitivity evaluation coefficient corresponding to the characteristic electricity consumption unit;

[0147] The step S333 further includes the following specific steps:

[0148] Step S3331: Mark multiple time periods in the historical electricity consumption data that are the same as the time range of the electricity consumption characteristic period as price sensitivity evaluation periods, and obtain multiple price sensitivity evaluation periods, and the electricity price corresponding to each price sensitivity evaluation period is different;

[0149] Step S3332: acquiring the unit electricity price of the characteristic electricity unit in each price sensitivity assessment period respectively, obtaining the unit electricity prices of multiple periods, and arranging the obtained unit electricity prices of the multiple periods in descending order according to the numerical values, to obtain a unit electricity price descending order queue;

[0150] Step S3333: In the descending order of the unit electricity price queue, select several time period unit electricity prices as characteristic unit electricity prices, and the price difference between every two consecutive characteristic unit electricity prices is the same, obtain the electricity price value for each characteristic unit electricity price, obtain multiple characteristic unit electricity price values, and name the obtained multiple characteristic unit electricity price values ​​from the first characteristic unit electricity price to the pth unit electricity price in order from small to large;

[0151] Step S3334: obtaining the power consumption of the characteristic power consumption unit in the price sensitivity assessment period of the first characteristic unit electricity price to obtain the first characteristic power consumption, obtaining the power consumption of the characteristic power consumption unit in the price sensitivity assessment period of the second characteristic unit electricity price to obtain the second characteristic power consumption, and so on, obtaining the power consumption of the characteristic power consumption unit in the price sensitivity assessment period of the pth characteristic unit electricity price to obtain the pth characteristic power consumption;

[0152] Step S3335: Figure 3 As shown, in the plane rectangular coordinate system, the characteristic electricity consumption is set as the coordinate y-axis, and the characteristic unit electricity price is set as the coordinate x-axis, and the price sensitivity analysis coordinate system is obtained;

[0153] Step S3336: In the price sensitivity analysis coordinate system, the first characteristic unit electricity price is used as the first horizontal coordinate, the first characteristic power consumption is used as the first vertical coordinate, and the first characteristic coordinate point is obtained; the second characteristic unit electricity price is used as the second horizontal coordinate, the second characteristic power consumption is used as the second vertical coordinate, and the first characteristic coordinate point is obtained; and so on, the pth characteristic unit electricity price is used as the pth horizontal coordinate, the pth characteristic power consumption is used as the pth vertical coordinate, and the pth characteristic coordinate point is obtained;

[0154] Step S3337: Mark the first characteristic coordinate point to the pth characteristic coordinate point in the price sensitivity analysis coordinate system respectively, connect the first characteristic coordinate point with the second characteristic coordinate point through the straight line tool to obtain the first characteristic polyline, connect the second characteristic coordinate point with the second characteristic coordinate point through the straight line tool to obtain the second characteristic polyline, repeat this process, connect the p-1th characteristic coordinate point with the pth characteristic coordinate point through the straight line tool to obtain the p-1th characteristic polyline;

[0155] Step S3338: Obtain the slopes of the first characteristic broken line to the p-1th characteristic broken line respectively, obtain p-1 broken line slopes, and calculate the average of the obtained p-1 broken line slopes to obtain the price sensitivity assessment coefficient corresponding to the characteristic electricity unit;

[0156] Step S334: Calculate the estimated power demand and estimated power supply in the power consumption characteristic period of the current year according to the power demand and power supply change trends in the power consumption characteristic period of the power company in the past five years, compare the estimated power demand and the estimated power supply, and when the estimated power demand is greater than or equal to the estimated power supply, use the value 0 to assign a parameter to the power supply index value, and when the estimated power demand is less than the estimated power supply, calculate the difference between the estimated power supply and the estimated power demand, take the absolute value of the obtained difference, and name the obtained difference as the power supply index value;

[0157] Step S335: defining the estimated power consumption during a time period, the price sensitivity evaluation coefficient, and the power supply index value as time period analysis data corresponding to the characteristic power consumption unit;

[0158] Step S34: acquiring time period analysis data of each first type electricity user in the characteristic electricity user unit to obtain a plurality of time period analysis data;

[0159] Step S35: defining the power consumption characteristic time period and the time period power consumption analysis data corresponding to the plurality of first type power consumption units as the time period power consumption analysis data;

[0160] Step S4: dividing multiple power consumption characteristic time periods into a first type of power consumption time period and a second type of power consumption time period according to the time period power consumption analysis data, obtaining power consumption characteristic time period division data, and performing power trading control on regional power consumption units according to the power consumption unit type division data and the power consumption characteristic time period division data;

[0161] Step S41: acquiring electricity consumption unit type classification data, and acquiring a plurality of first-type electricity consumption units and second-type electricity consumption units according to the electricity consumption unit type classification data;

[0162] Step S42: acquiring time period electricity consumption analysis data, and acquiring time period analysis data corresponding to the electricity consumption characteristic time period and each first type electricity consumption unit respectively according to the time period electricity consumption analysis data;

[0163] Step S43: obtaining the estimated power consumption, price sensitivity evaluation coefficient and power supply index value of each time period according to the time period analysis data;

[0164] Step S44: Calculate the power consumption time period type division coefficient by using the estimated power consumption time period, the price sensitivity evaluation coefficient and the power supply index value;

[0165] The specific formula for calculating the electricity consumption time period classification coefficient is as follows:

[0166]

[0167] Among them, Sdh is the electricity consumption period type classification coefficient, Mgx is the price sensitivity assessment coefficient, Gdz is the power supply index value, and Ydl is the estimated power consumption during the period;

[0168] Step S45: respectively acquiring the power consumption period type division coefficient corresponding to each first type power consumption unit in the power consumption characteristic period, to obtain a plurality of power consumption period type division coefficients;

[0169] Step S46: obtaining a power consumption period type division coefficient threshold, comparing the multiple power consumption period type division coefficients with the power consumption period type division coefficient threshold, dividing the power consumption characteristic period into a first type power consumption period and a second type power consumption period according to the numerical comparison result, and obtaining power consumption characteristic period division data;

[0170] The step S46 further includes the following specific steps:

[0171] Step S461: setting a time period estimated power consumption threshold, a price sensitivity assessment coefficient threshold, and a power supply index threshold respectively;

[0172] Step S462: Calculate the power consumption time period type classification coefficient threshold by using the time period estimated power consumption threshold, the price sensitivity evaluation coefficient threshold and the power supply index threshold;

[0173] The power consumption period type division coefficient threshold is calculated, and the specific formula is as follows:

[0174]

[0175] Among them, Sdh is the threshold of the electricity consumption period type classification coefficient, Mgx is the threshold of the price sensitivity assessment coefficient, Ydl is the threshold of the estimated electricity consumption during the period, and Gdz is the threshold of the power supply index;

[0176] Step S463: if the power consumption period type division coefficient is greater than or equal to the power consumption period type division coefficient threshold, the power consumption feature period is divided into the first type of power consumption period;

[0177] Step S464: if the power consumption period type division coefficient is less than the power consumption period type division coefficient threshold, the power consumption characteristic period is divided into the second type power consumption period;

[0178] Step S47: Controlling power trading of regional power users according to the power user type classification results and power consumption characteristic time period classification results;

[0179] The step S47 further includes the following specific steps:

[0180] Step S471: When the regional power user is a first type power user and the power consumption characteristic period is a first type power consumption period, the power company stops conducting power transactions with the regional power user during the power consumption characteristic period;

[0181] Step S472: When the regional power user is a first type power user and the power consumption characteristic period is a second type power consumption period, the power company increases the power transaction price for the regional power user;

[0182] Step S473: When the regional power user is a second type of power user, the power company conducts power transactions with the regional power user as normal.

[0183] In this application, if corresponding calculation formulas appear, the above calculation formulas are all dimensionless and take their numerical calculations. The weight coefficients, proportional coefficients and other coefficients in the formulas are set to a result value obtained by quantifying each parameter. The size of the weight coefficient and the proportional coefficient can be determined as long as it does not affect the proportional relationship between the parameter and the result value.

[0184] Embodiment 2, as Figure 2 As shown, based on the same inventive concept, this embodiment also provides an electric power trading system based on multi-party secure computing and federated learning, including:

[0185] The data acquisition module is used to acquire the power trading area and respectively acquire the power unit historical monitoring data corresponding to each power unit in the power trading area; the power unit historical monitoring data includes: abnormal power cycle number ratio, cycle average power consumption deviation and cycle power consumption deviation variance;

[0186] A unit analysis module, used to classify regional power users in the power trading area into first type power users and second type power users according to historical monitoring data of power users;

[0187] The time period analysis module is used to mark a power consumption characteristic time period in the future power consumption time period of the first type of power consumption unit according to the power consumption unit type classification result, and obtain the time period power consumption analysis data by performing power analysis on the power consumption characteristic time period to obtain the time period estimated power consumption, price sensitivity evaluation coefficient and power supply index value corresponding to each first type of power consumption unit in the power consumption characteristic time period;

[0188] The power trading module is used to divide multiple power consumption characteristic time periods into the first type of power consumption time period and the second type of power consumption time period according to the power consumption analysis data of the time period, and to control the power trading of regional power consumption units according to the power consumption unit type division results and the power consumption characteristic time period division results.

[0189] The data acquisition module, unit analysis module, time period analysis module and power trading module are respectively connected to the server, and the server controls the data acquisition module, unit analysis module, time period analysis module and power trading module respectively.

[0190] The data acquisition module acquires the power trading area covered by the power company, acquires each power user in the power trading area, obtains multiple regional power users, and randomly selects one power user from the multiple regional power users as a sample power user;

[0191] It should be noted here that:

[0192] In this application, the power trading area referred to here is the power supply area covered by the power company;

[0193] In this application, the power transaction involved here specifically refers to the power transaction between the power company and the power user, that is, the power company transmits power resources to the power user and collects electricity fees;

[0194] In this application, the power trading area involved here is specifically the household power consumption area, and the regional power consumption unit involved here is specifically the household;

[0195] Conduct power transaction monitoring on sample power users to obtain historical monitoring data corresponding to the sample power users;

[0196] The details are as follows:

[0197] Obtain historical electricity consumption data corresponding to the sample electricity consumption unit, mark several historical electricity consumption monitoring cycles with a time length as the characteristic power monitoring time in the historical electricity consumption data, and the time interval between each two consecutive historical electricity consumption monitoring cycles is the same;

[0198] It should be noted here that:

[0199] The time value corresponding to the characteristic power monitoring duration involved here is specifically 24 hours;

[0200] The obtained multiple historical power consumption monitoring cycles are named as the first historical power consumption monitoring cycle to the i-th historical power consumption monitoring cycle respectively;

[0201] It should be noted here that:

[0202] In this application, i is the quantity value corresponding to the historical electricity consumption monitoring period, and i is an integer greater than 0;

[0203] Perform unit power analysis on the first historical power consumption monitoring period to the i-th historical power consumption monitoring period to obtain the ratio of the number of abnormal power periods;

[0204] The details are as follows:

[0205] In the first historical power consumption monitoring cycle, a number of power monitoring time points are randomly selected, and the obtained power monitoring time points are named as the first power monitoring time point to the jth power monitoring time point respectively;

[0206] It should be noted here that:

[0207] In this application, j is the quantity value corresponding to the power monitoring time point, and j is an integer greater than 0;

[0208] Respectively obtain the power consumption values ​​corresponding to the sample power consumption units from the first power monitoring time point to the jth power monitoring time point, and obtain the first power consumption value to the jth power consumption value;

[0209] Obtain the reference power values ​​corresponding to the first power monitoring time point to the jth power monitoring time point in the power trading area, and obtain the first reference power value to the jth reference power value;

[0210] It should be noted here that:

[0211] In the present application, the reference power consumption value is specifically the average power consumption value of the power trading area at the corresponding power monitoring time point, for example: the first reference power consumption value is the average power consumption value of all regional power users in the power trading area at the first power monitoring time point;

[0212] The first power consumption value to the jth power consumption value and the first reference power consumption value to the jth reference power consumption value are calculated to obtain the period power consumption deviation corresponding to the first historical power consumption monitoring period;

[0213] The periodic power deviation corresponding to the first historical power monitoring period is calculated, and the specific formula is as follows:

[0214]

[0215] Wherein, Wpc is the periodic power deviation corresponding to the first historical power monitoring period, Wsl1 to Wslj are respectively the first power value to the jth power value, Wsj1 to Wsjj are respectively the first reference power value to the jth reference power value, and j is the quantity value corresponding to the power monitoring time point;

[0216] Repeat the process of obtaining the periodic power consumption deviation corresponding to the first historical power consumption monitoring cycle, respectively obtain the periodic power consumption deviations corresponding to the second historical power consumption monitoring cycle to the i-th historical power consumption monitoring cycle, and obtain multiple periodic power consumption deviations;

[0217] Obtain a periodic power consumption deviation threshold, and compare multiple periodic power consumption deviations with multiple periodic power consumption deviation thresholds respectively; when the periodic power consumption deviation is greater than or equal to the periodic power consumption deviation threshold, the corresponding historical power consumption monitoring period is marked as an abnormal power monitoring period; when the periodic power consumption deviation is less than the periodic power consumption deviation threshold, the corresponding historical power consumption monitoring period is marked as a normal power monitoring period;

[0218] Perform quantity statistics on the first historical power consumption monitoring cycle to the i-th historical power consumption monitoring cycle to obtain the number of abnormal power monitoring cycles, calculate the ratio of the number of abnormal power monitoring cycles to i, and obtain the number ratio of abnormal power cycles;

[0219] The periodic power consumption of the sample power consumption units from the first historical power consumption monitoring period to the i-th historical power consumption monitoring period is obtained respectively, and the power consumption from the first historical period to the i-th historical period is obtained;

[0220] Respectively obtain the average power consumption per unit period of all regions corresponding to the power trading area in the first historical power consumption monitoring period to the i-th historical power consumption monitoring period, and obtain the benchmark power consumption of the first historical period to the i-th historical period;

[0221] Calculate the difference between the electricity consumption of the first historical period and the benchmark electricity consumption of the first historical period to obtain the electricity consumption deviation of the first period, calculate the difference between the electricity consumption of the second historical period and the benchmark electricity consumption of the second historical period to obtain the electricity consumption deviation of the second period, repeat this process, and obtain the electricity consumption deviation of the third period to the electricity consumption deviation of the i-th period respectively;

[0222] The average of the power consumption deviation from the first cycle to the power consumption deviation from the i-th cycle is calculated to obtain the average power consumption deviation of the cycle;

[0223] Calculate the variance of the first cycle power consumption deviation to the i-th cycle power consumption deviation to obtain the variance of the cycle power consumption deviation;

[0224] The abnormal power cycle number ratio, cycle average power consumption deviation and cycle power consumption deviation variance corresponding to the sample power consumption units are defined as historical monitoring data;

[0225] The historical monitoring data corresponding to each regional power consumption unit are respectively acquired to obtain a plurality of historical monitoring data;

[0226] A plurality of historical monitoring data are defined as historical monitoring data of power consumption units;

[0227] The data acquisition module acquires the historical monitoring data of the power consumption unit and transmits it to the unit analysis module;

[0228] The unit analysis module divides the regional power users in the power trading area into first-type power users and second-type power users according to the historical monitoring data of the power users, and obtains the power user type division data;

[0229] Obtain historical monitoring data of power consumption units, and obtain historical monitoring data corresponding to each power consumption unit in each region according to the historical monitoring data of power consumption units;

[0230] Performing power consumption data analysis on the sample power consumption units according to the historical monitoring data corresponding to the sample power consumption units, and obtaining power data abnormality analysis coefficients corresponding to the sample power consumption units;

[0231] The details are as follows:

[0232] According to the historical monitoring data corresponding to the sample electricity consumption units, the abnormal power cycle number ratio, the cycle average power consumption deviation and the cycle power consumption deviation variance are obtained respectively;

[0233] The power data abnormality analysis coefficient is obtained by calculating the abnormal power cycle number ratio, cycle average power consumption deviation and cycle power consumption deviation variance;

[0234] The power data abnormality analysis coefficient is calculated, and the specific formula is as follows:

[0235] Dyx=Zyb+Zpp*(1+Zfc)+Zpp;

[0236] Among them, Dyx is the power data abnormality analysis coefficient, Zyb is the abnormal power cycle number ratio, Zpp is the cycle average power consumption deviation, and Zfc is the cycle power consumption deviation variance;

[0237] Repeat the process of obtaining the power data abnormality analysis coefficient corresponding to the sample power consumption unit, respectively obtain the power data abnormality analysis coefficient corresponding to each regional power consumption unit, and obtain multiple power data abnormality analysis coefficients;

[0238] The power data abnormality analysis coefficient corresponding to each regional power consumption unit is obtained respectively to obtain a plurality of power data abnormality analysis coefficients, and a power data abnormality analysis coefficient threshold is obtained. Each power data abnormality analysis coefficient is numerically compared with the power data abnormality analysis coefficient threshold, and the regional power consumption units are divided into a first type of power consumption unit and a second type of power consumption unit according to the numerical comparison result to obtain power consumption unit type division data;

[0239] The details are as follows:

[0240] Obtaining respectively the abnormal power cycle quantity ratio threshold, the cycle average power consumption deviation threshold and the cycle power consumption deviation variance threshold;

[0241] It should be noted here that:

[0242] In the present application, the abnormal power cycle number ratio threshold, the cycle average power consumption deviation threshold, and the cycle power consumption deviation variance threshold are all the maximum abnormal power cycle number ratio, the maximum cycle average power consumption deviation, and the maximum cycle power consumption deviation variance corresponding to the second type of power consumption unit;

[0243] The power data abnormality analysis coefficient threshold is obtained by calculating the abnormal power cycle number ratio threshold, the cycle average power consumption deviation threshold and the cycle power consumption deviation variance threshold;

[0244] The power data abnormality analysis coefficient threshold is calculated, and the specific formula is as follows:

[0245] Dyxy=Zyby+Zppy*(1+Zfcy)+Zppy;

[0246] Among them, Dyxy is the power data abnormality analysis coefficient threshold, Zyby is the abnormal power cycle number ratio threshold, Zppy is the cycle average power consumption deviation threshold, and Zfcy is the cycle power consumption deviation variance threshold;

[0247] When the power data abnormality analysis coefficient is greater than or equal to the power data abnormality analysis coefficient threshold, the corresponding regional power consumption unit is classified as a first type of power consumption unit;

[0248] When the power data abnormality analysis coefficient is less than the power data abnormality analysis coefficient threshold, the corresponding regional power consumption unit is classified as a second type of power consumption unit;

[0249] It should be noted here that:

[0250] In this application, the first type of power users are power users in the region where power trading is abnormal, and their power trading methods need to be further analyzed and determined. The second type of power users are power users in the region where power trading is normal, and they can continue to trade power using the current power trading methods.

[0251] The unit analysis module acquires the data on the classification of power consumption units and transmits it to the time period analysis module and the power trading module;

[0252] The time period analysis module marks a power consumption characteristic time period in the future power consumption period of the first type of power consumption unit according to the power consumption unit type classification data, and obtains the time period estimated power consumption, price sensitivity evaluation coefficient and power supply index value corresponding to each first type of power consumption unit by performing power analysis on the power consumption characteristic time period, thereby obtaining the time period power consumption analysis data;

[0253] Acquire power unit type classification data, acquire each first type power unit in the power trading area according to the power unit type classification data, and randomly select one power unit from the acquired multiple first type power units as a characteristic power unit;

[0254] Mark the time point corresponding to the current moment as the first power consumption time point, mark the time point corresponding to a characteristic power consumption duration after the first power consumption time point as the second power consumption time point, and mark the period between the first power consumption time point and the second power consumption time point as the power consumption characteristic period;

[0255] It should be noted here that:

[0256] In the present application, as the time point corresponding to the current time changes, the first power consumption time point and the second power consumption time point also change accordingly, thereby realizing the dynamic update of the power consumption characteristic time period;

[0257] Performing time period power analysis on characteristic power consumption units in characteristic power consumption time periods to obtain time period analysis data corresponding to the characteristic power consumption units;

[0258] The details are as follows:

[0259] Obtain historical electricity consumption data corresponding to the characteristic electricity consumption unit, mark an electricity consumption evaluation period in the historical electricity consumption data, and obtain multiple time periods in the electricity consumption evaluation period that are the same as the time range of the electricity consumption characteristic period as the electricity consumption evaluation period, to obtain multiple electricity consumption evaluation periods;

[0260] It should be noted here that:

[0261] The electricity consumption assessment period here is specifically one month before the first electricity consumption time point;

[0262] The electricity consumption evaluation period involved here is specifically the period in the month before the first electricity consumption time point, which is the same as the time range corresponding to the electricity consumption characteristic period;

[0263] For example, if the time range corresponding to the electricity consumption characteristic period is 9:00-9:15 on November 1, the time range corresponding to the electricity consumption assessment period can be 9:00-9:15 on October 20, 9:00-9:15 on October 22, and 9:00-9:15 on October 23;

[0264] The power consumption of the characteristic power consumption unit in each power consumption evaluation period is obtained respectively, and the power consumption of multiple periods is obtained, and the average of the obtained power consumption of multiple periods is calculated to obtain the estimated power consumption of the characteristic power consumption unit in the power consumption characteristic period;

[0265] Obtain the price sensitivity assessment coefficient corresponding to the characteristic electricity consumption unit;

[0266] The details are as follows:

[0267] Mark multiple time periods in the historical electricity consumption data that are the same as the time range of the electricity consumption characteristic period as price sensitivity evaluation periods, and obtain multiple price sensitivity evaluation periods, and the electricity price corresponding to each price sensitivity evaluation period is different;

[0268] The unit electricity price of the characteristic electricity user in each price sensitivity assessment period is obtained respectively to obtain the unit electricity prices of multiple periods, and the obtained unit electricity prices of the multiple periods are arranged in descending order according to the numerical values ​​to obtain a unit electricity price descending order queue;

[0269] In the descending order of the unit electricity price queue, select several time period unit electricity prices as characteristic unit electricity prices, and the price difference between every two consecutive characteristic unit electricity prices is the same;

[0270] It should be noted here that:

[0271] In the present application, the difference in the amount of electricity price between each two consecutive characteristic unit electricity prices is the same, which can be specifically understood as a plurality of characteristic unit electricity prices can form an arithmetic progression. For example, assuming that three characteristic unit electricity prices are selected, which are 0.5, 0.7 and 0.9 respectively, these three characteristic unit electricity prices can form an arithmetic progression with a first term of 0.5 and a tolerance of 0.2.

[0272] Obtaining the electricity price value for each characteristic unit electricity price respectively, obtaining multiple characteristic unit electricity price values, and naming the obtained multiple characteristic unit electricity price values ​​in order from small to large as the first characteristic unit electricity price to the pth unit electricity price;

[0273] It should be noted here that:

[0274] In this application, p is the quantity value corresponding to the characteristic unit electricity price, and p is an integer greater than 0;

[0275] Obtain the power consumption of the characteristic power unit in the price sensitivity assessment period of the first characteristic unit electricity price to obtain the first characteristic power consumption, obtain the power consumption of the characteristic power unit in the price sensitivity assessment period of the second characteristic unit electricity price to obtain the second characteristic power consumption, and so on, obtain the power consumption of the characteristic power unit in the price sensitivity assessment period of the pth characteristic unit electricity price to obtain the pth characteristic power consumption;

[0276] In the plane rectangular coordinate system, the characteristic electricity consumption is set as the coordinate y-axis, and the characteristic unit electricity price is set as the coordinate x-axis to obtain the price sensitivity analysis coordinate system;

[0277] In the price sensitivity analysis coordinate system, the first characteristic unit electricity price is taken as the first horizontal coordinate, the first characteristic power consumption is taken as the first vertical coordinate, and the first characteristic coordinate point is obtained; the second characteristic unit electricity price is taken as the second horizontal coordinate, the second characteristic power consumption is taken as the second vertical coordinate, and the first characteristic coordinate point is obtained; and so on, the pth characteristic unit electricity price is taken as the pth horizontal coordinate, the pth characteristic power consumption is taken as the pth vertical coordinate, and the pth characteristic coordinate point is obtained;

[0278] Mark the first characteristic coordinate point to the pth characteristic coordinate point in the price sensitivity analysis coordinate system respectively, connect the first characteristic coordinate point with the second characteristic coordinate point through the straight line tool to obtain the first characteristic polyline, connect the second characteristic coordinate point with the second characteristic coordinate point through the straight line tool to obtain the second characteristic polyline, repeat this process, connect the p-1th characteristic coordinate point with the pth characteristic coordinate point through the straight line tool to obtain the p-1th characteristic polyline;

[0279] Obtain the slopes of the first characteristic broken line to the p-1th characteristic broken line respectively, obtain p-1 broken line slopes, and calculate the average of the obtained p-1 broken line slopes to obtain the price sensitivity assessment coefficient corresponding to the characteristic electricity unit;

[0280] Obtain the estimated power demand and estimated power supply of the power company during the power consumption characteristic period, compare the estimated power demand and the estimated power supply, and when the estimated power demand is greater than or equal to the estimated power supply, use the value 0 to assign a parameter to the power supply index value; when the estimated power demand is less than the estimated power supply, calculate the difference between the estimated power supply and the estimated power demand, take the absolute value of the obtained difference, and name the obtained difference as the power supply index value;

[0281] It should be noted here that:

[0282] The estimated power demand involved here is the estimated value of the power required in the power trading area, and the estimated power supply involved here is specifically the estimated value of the power that the power company can provide for the power trading area. The estimated power demand and estimated power supply involved here are both provided by the power company's existing power forecasting model;

[0283] The estimated power consumption during a period, the price sensitivity evaluation coefficient and the power supply index value are defined as the period analysis data corresponding to the characteristic power consumption unit;

[0284] Acquire the time period analysis data of each first type electricity user in the characteristic electricity user to obtain a plurality of time period analysis data;

[0285] The characteristic power consumption period and the power consumption analysis data of multiple periods are defined as the power consumption analysis data of the period;

[0286] The time period analysis module acquires the time period power consumption analysis data and transmits it to the power trading module;

[0287] The power trading module divides multiple power consumption characteristic time periods into a first type of power consumption time period and a second type of power consumption time period according to the power consumption analysis data of the time period, obtains power consumption characteristic time period division data, and performs power trading control on regional power consumption units according to the power consumption unit type division data and the power consumption characteristic time period division data;

[0288] Acquire electricity user unit type classification data, and acquire a plurality of first-type electricity users and second-type electricity users according to the electricity user unit type classification data;

[0289] Acquire time period electricity consumption analysis data, and acquire time period analysis data corresponding to the electricity consumption characteristic time period and each first type electricity consumption unit according to the time period electricity consumption analysis data;

[0290] According to the time period analysis data, the estimated power consumption, price sensitivity evaluation coefficient and power supply index value of each time period are obtained respectively;

[0291] The power consumption time period classification coefficient is obtained by calculating the estimated power consumption, price sensitivity evaluation coefficient and power supply index value;

[0292] The specific formula for calculating the electricity consumption time period classification coefficient is as follows:

[0293]

[0294] Among them, Sdh is the electricity consumption period type classification coefficient, Mgx is the price sensitivity assessment coefficient, Gdz is the power supply index value, and Ydl is the estimated power consumption during the period;

[0295] Obtaining the electricity consumption period type division coefficient corresponding to each first type electricity consumption unit in the electricity consumption characteristic period respectively, to obtain a plurality of electricity consumption period type division coefficients;

[0296] Obtaining a power consumption period type division coefficient threshold, comparing multiple power consumption period type division coefficients with the power consumption period type division coefficient threshold, dividing the power consumption characteristic period into a first type power consumption period and a second type power consumption period according to the numerical comparison result, and obtaining power consumption characteristic period division data;

[0297] The power consumption period type division coefficient threshold is obtained as follows:

[0298] Obtain the estimated power consumption threshold for the time period, the price sensitivity assessment coefficient threshold, and the power supply index threshold respectively;

[0299] It should be noted here that:

[0300] In the present application, the time period estimated power consumption threshold, the price sensitivity assessment coefficient threshold, and the power supply index threshold are all the maximum time period estimated power consumption, the maximum price sensitivity assessment coefficient, and the maximum power supply index value corresponding to the second type of power consumption time period;

[0301] The power consumption period type classification coefficient threshold is obtained by calculating the power consumption period estimated power consumption threshold, price sensitivity assessment coefficient threshold and power supply index threshold;

[0302] The power consumption period type division coefficient threshold is calculated, and the specific formula is as follows:

[0303]

[0304] Among them, Sdh is the threshold of the electricity consumption period type classification coefficient, Mgx is the threshold of the price sensitivity assessment coefficient, Ydl is the threshold of the estimated electricity consumption during the period, and Gdz is the threshold of the power supply index;

[0305] The numerical comparison process is as follows:

[0306] If the power consumption period type division coefficient is greater than or equal to the power consumption period type division coefficient threshold, the power consumption characteristic period is divided into the first type of power consumption period;

[0307] If the power consumption period type division coefficient is less than the power consumption period type division coefficient threshold, the power consumption characteristic period is divided into the second type of power consumption period;

[0308] Control power trading of regional power users based on data on power user types and power consumption feature time periods;

[0309] The details are as follows:

[0310] When the regional electricity user is a first type of electricity user and the electricity consumption characteristic period is a first type of electricity consumption period, the power company stops conducting electricity transactions with the regional electricity user during the electricity consumption characteristic period;

[0311] When the regional electricity user is a first type of electricity user and the characteristic period of electricity consumption is a second type of electricity consumption period, the power company increases the electricity transaction price for the regional electricity user;

[0312] When the regional electricity user is a second type of electricity user, the power company normally conducts electricity transactions with the regional electricity user.

[0313] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A power trading method based on multi-party secure computing and federated learning, characterized in that: include: S1: Obtain the power trading area, and respectively obtain the power user historical monitoring data corresponding to each regional power user in the power trading area; The historical monitoring data of the power consumption unit includes: abnormal power cycle number ratio, cycle average power consumption deviation and cycle power consumption deviation variance; S2: Divide the regional power users in the power trading area into first-type power users and second-type power users according to the historical monitoring data of the power users; S3: According to the classification result of the power consumption unit type, a power consumption characteristic period is marked in the future power consumption period of the first type of power consumption unit, and the power consumption characteristic period is analyzed to obtain the estimated power consumption, price sensitivity evaluation coefficient and power supply index value of each first type of power consumption unit in the power consumption characteristic period, so as to obtain the power consumption analysis data of the period; S4: Divide a plurality of power consumption characteristic time periods into a first type of power consumption time period and a second type of power consumption time period according to the power consumption analysis data during the time period, and perform power trading control on regional power consumption units according to the power consumption unit type division result and the power consumption characteristic time period division result.

2. The power trading method based on multi-party secure computing and federated learning according to claim 1 is characterized in that: The step S1 comprises: S11: Obtain the power trading area covered by the power company and each regional power user in the power trading area, and randomly select one power user from each regional power user as a sample power user; S12: monitoring the power transaction of the sample power users to obtain historical monitoring data of the sample power users; S13: Based on the historical monitoring data of the sample electricity consumption units, respectively obtain the historical monitoring data corresponding to each regional electricity consumption unit; S14: defining the historical monitoring data corresponding to the power consumption unit in each area as the power consumption unit historical monitoring data.

3. The power trading method based on multi-party secure computing and federated learning according to claim 2 is characterized in that: The step S12 comprises: S121: Obtain historical electricity consumption data corresponding to a sample electricity consumption unit, and divide the historical electricity consumption data into a number of historical electricity consumption monitoring periods based on a preset monitoring time period; the historical electricity consumption data includes electricity consumption and electricity power; S122: Assume that each historical power consumption monitoring period is from the first historical power consumption monitoring period to the i-th historical power consumption monitoring period; S123: Perform unit power analysis on all historical power monitoring cycles of the sample power-consuming units to obtain the ratio of the number of abnormal power cycles; specifically, the following includes: In the first historical power consumption monitoring cycle, a number of power monitoring time points are randomly selected, and the obtained power monitoring time points are named as the first power monitoring time point to the jth power monitoring time point respectively; Respectively obtain the power consumption values ​​corresponding to the sample power consumption units from the first power monitoring time point to the jth power monitoring time point, and obtain the first power consumption value to the jth power consumption value; Obtain the reference power values ​​corresponding to the power trading area from the first power monitoring time point to the jth power monitoring time point, and obtain the first reference power value to the jth reference power value; The periodic power deviation corresponding to the first historical power monitoring period is obtained by calculating the first power consumption value to the jth power consumption value and the first reference power consumption value to the jth reference power consumption value. The specific formula is as follows: Wherein, Wpc is the periodic power deviation corresponding to the first historical power monitoring period, Wsl1 to Wslj are respectively the first power value to the jth power value, Wsj1 to Wsjj are respectively the first reference power value to the jth reference power value, and j is the number of power monitoring time points; Respectively obtain the periodic power deviations corresponding to the second historical power monitoring period to the i-th historical power monitoring period to obtain the corresponding periodic power deviations; Set a periodic power consumption deviation threshold, compare the power consumption deviation of each period with multiple periodic power consumption deviation thresholds, and when a periodic power consumption deviation is greater than or equal to the periodic power consumption deviation threshold, mark the corresponding historical power consumption monitoring period as an abnormal power monitoring period; when a periodic power consumption deviation is less than the periodic power consumption deviation threshold, mark the corresponding historical power consumption monitoring period as a normal power monitoring period; Counting the number of abnormal power monitoring cycles from the first historical power consumption monitoring cycle to the i-th historical power consumption monitoring cycle to obtain a number value of abnormal power monitoring cycles, and calculating a ratio of the number value of abnormal power monitoring cycles to i to obtain a number ratio of abnormal power cycles; S124: Obtaining the power consumption of the sample power consumption unit from the first historical period to the i-th historical period; S125: Obtain the average power consumption per unit period of all regions in the power trading area from the first historical power consumption monitoring period to the i-th historical power consumption monitoring period, and obtain the benchmark power consumption from the first historical period to the i-th historical period; S126: Calculate the difference between the power consumption of each historical period of the sample power consumption unit and the corresponding historical period benchmark power consumption, and obtain the power consumption deviation from the first historical period to the i-th historical period; S127: Calculate the average of the power consumption deviation from the first historical period to the power consumption deviation from the i-th historical period to obtain the period average power consumption deviation; S128: Calculate the variance of the first historical period power consumption deviation to the i-th historical period power consumption deviation to obtain the period power consumption deviation variance; S129: The abnormal power cycle number ratio, cycle average power consumption deviation, and cycle power consumption deviation variance corresponding to the sample power consumption units are defined as historical monitoring data.

4. The power trading method based on multi-party secure computing and federated learning according to claim 3 is characterized in that: The step S2 comprises: S21: Obtaining historical monitoring data corresponding to each regional electricity user; S22: Analyze the power consumption data according to the historical monitoring data of the sample power consumption units to obtain the power data abnormality analysis coefficients of the sample power consumption units, and analyze the power consumption data to obtain the power data abnormality analysis coefficients corresponding to the power consumption units in each region; S23: Based on the calculation method of the power data anomaly analysis coefficient of the sample power units, the power data anomaly analysis coefficient corresponding to each regional power unit is calculated respectively to obtain the corresponding power data anomaly analysis coefficient; a power data anomaly analysis coefficient threshold corresponding to each regional power unit is set, and each power data anomaly analysis coefficient is numerically compared with the power data anomaly analysis coefficient threshold, and the regional power units are divided into first type power units and second type power units according to the numerical comparison results to obtain power unit type classification data.

5. The power trading method based on multi-party secure computing and federated learning according to claim 4 is characterized in that: The step S22 comprises: S221: Obtaining the abnormal power cycle number ratio, cycle average power consumption deviation, and cycle power consumption deviation variance of the sample power consumption unit according to the historical monitoring data corresponding to the sample power consumption unit; S222: Calculate the power data abnormality analysis coefficient based on the abnormal power cycle number ratio, cycle average power consumption deviation, and cycle power consumption deviation variance of the sample power consumption unit: Dyx=Zyb+Zpp*(1+Zfc)+Zpp Among them, Dyx is the power data abnormality analysis coefficient, Zyb is the ratio of the number of abnormal power cycles, Zpp is the cycle average power consumption deviation, and Zfc is the variance of the cycle power consumption deviation.

6. The power trading method based on multi-party secure computing and federated learning according to claim 5 is characterized in that: The step S23 comprises: Step S231: setting the abnormal power cycle number ratio threshold, cycle average power consumption deviation threshold and cycle power consumption deviation variance threshold of each regional power consumption unit respectively; Step S232: Calculate the power data abnormality analysis coefficient threshold by using the abnormal power cycle number ratio threshold, the cycle average power consumption deviation threshold, and the cycle power consumption deviation variance threshold; Step S233: when the power data abnormality analysis coefficient is greater than or equal to the power data abnormality analysis coefficient threshold, the corresponding regional power consumption unit is classified as a first type of power consumption unit; Step S234: when the power data abnormality analysis coefficient is less than the power data abnormality analysis coefficient threshold, the corresponding regional power consumption unit is classified as a second type of power consumption unit.

7. The power trading method based on multi-party secure computing and federated learning according to claim 6 is characterized in that: The step S3 comprises: S31: According to the classification result of the electricity consumption unit type, randomly select one electricity consumption unit from all the first type electricity consumption units as a characteristic electricity consumption unit; S32: Preset a time length as a characteristic power usage duration, mark the time point corresponding to the current moment as a first power usage time point, mark a time point corresponding to a characteristic power usage duration after the first power usage time point as a second power usage time point, and mark the period between the first power usage time point and the second power usage time point as a power usage characteristic period; S33: Performing time period power analysis on the characteristic time period of the characteristic power consumption unit to obtain time period analysis data corresponding to the characteristic power consumption unit; S34: based on the time period analysis data corresponding to the characteristic electricity consumption unit, respectively acquiring the time period analysis data of each first type electricity consumption unit in the characteristic electricity consumption unit, to obtain a plurality of time period analysis data; S35: defining the power consumption characteristic time period and the time period power consumption analysis data corresponding to the plurality of first type power consumption units as the time period power consumption analysis data.

8. The power trading method based on multi-party secure computing and federated learning according to claim 7 is characterized in that: The step S33 comprises: S331: Marking a power consumption evaluation period in the historical power consumption data corresponding to the characteristic power consumption unit, and obtaining multiple time periods in the power consumption evaluation period that are the same as the power consumption characteristic time period as the power consumption evaluation period, to obtain multiple power consumption evaluation periods; S332: respectively obtaining the power consumption of the characteristic power consumption unit in each power consumption evaluation period, obtaining the power consumption of multiple periods, and averaging the obtained power consumption of the multiple periods to obtain the estimated power consumption of the characteristic power consumption unit in the period corresponding to the power consumption characteristic period; S333: Obtaining the price sensitivity evaluation coefficient corresponding to the characteristic electricity consumption unit; specifically including: Mark multiple time periods in the historical electricity consumption data that are the same as the time range of the electricity consumption characteristic period as price sensitivity evaluation periods, and obtain multiple price sensitivity evaluation periods, and the electricity price corresponding to each price sensitivity evaluation period is different; Obtain the unit electricity price of the characteristic electricity user in each price sensitivity assessment period, and sort the obtained unit electricity prices in multiple periods in descending order according to the numerical value to obtain a unit electricity price descending sorting queue; In the descending order of the unit electricity price queue, several time period unit electricity prices are selected as characteristic unit electricity prices, and the price difference between two consecutive characteristic unit electricity prices is constrained to be the same, and the electricity price value is obtained for each characteristic unit electricity price to obtain multiple characteristic unit electricity price values, and the obtained multiple characteristic unit electricity price values ​​are named from the first characteristic unit electricity price to the pth unit electricity price in order from small to large; Obtain the power consumption of the characteristic power consumption unit in the price sensitivity assessment period from the first characteristic unit electricity price to the pth characteristic unit electricity price, and obtain the first characteristic power consumption to the pth characteristic power consumption; In the plane rectangular coordinate system, the characteristic electricity consumption is set as the coordinate y-axis, and the characteristic unit electricity price is set as the coordinate x-axis to obtain the price sensitivity analysis coordinate system; In the price sensitivity analysis coordinate system, the first characteristic unit electricity price to the pth characteristic unit electricity price are used as the horizontal coordinate, and the first characteristic power consumption to the pth characteristic power consumption are used as the vertical coordinate to obtain the first characteristic coordinate point to the pth characteristic coordinate point; Mark the first characteristic coordinate point to the pth characteristic coordinate point in the price sensitivity analysis coordinate system respectively, and connect each adjacent characteristic coordinate point to obtain p-1 characteristic polylines; Obtain the slopes of all p-1 characteristic broken lines, obtain p-1 broken line slopes, and average the obtained p-1 broken line slopes to obtain the price sensitivity assessment coefficient corresponding to the characteristic electricity unit; S334: Obtain the estimated power demand and estimated power supply of the power company during the power consumption characteristic period, set the power supply index value as a parameter, compare the estimated power demand and the estimated power supply, when the estimated power demand is greater than or equal to the estimated power supply, the power supply index value is 0, when the estimated power demand is less than the estimated power supply, the power supply index value is the absolute value of the difference between the estimated power supply and the estimated power demand; S335: The estimated power consumption during a time period, the price sensitivity evaluation coefficient, and the power supply index value are defined as the time period analysis data corresponding to the characteristic power consumption unit.

9. The power trading method based on multi-party secure computing and federated learning according to claim 8 is characterized in that: The step S4 comprises: S41: Calculate the corresponding electricity consumption period type division coefficient according to the estimated electricity consumption, price sensitivity evaluation coefficient and power supply index value of each first type electricity user unit during the period: Among them, Sdh is the electricity consumption period type classification coefficient, Mgx is the price sensitivity assessment coefficient, Gdz is the power supply index value, and Ydl is the estimated power consumption during the period; S42: Obtaining a power consumption period type division coefficient threshold, comparing the values ​​of each power consumption period type division coefficient with the power consumption period type division coefficient threshold, dividing the power consumption feature period into a first type power consumption period and a second type power consumption period according to the numerical comparison result, and obtaining the power consumption feature period division result; including: S421: setting a time period estimated power consumption threshold, a price sensitivity assessment coefficient threshold, and a power supply index threshold respectively; S422: Based on the calculation method of the electricity consumption period classification system, the estimated electricity consumption threshold, the price sensitivity evaluation coefficient threshold and the power supply index threshold are calculated to obtain the electricity consumption period classification coefficient threshold; S423: if the power consumption period type division coefficient is greater than or equal to the power consumption period type division coefficient threshold, the power consumption characteristic period is divided into a first type of power consumption period; S424: if the power consumption period type division coefficient is less than the power consumption period type division coefficient threshold, the power consumption characteristic period is divided into a second type of power consumption period; S43: Controlling power trading of regional power users according to the results of power user type classification and power consumption characteristic time period classification; specifically including: When the regional electricity user is a first type electricity user and the electricity consumption characteristic period is a first type electricity consumption period, the power company stops conducting electricity transactions with the regional electricity user during the electricity consumption characteristic period; When the regional electricity user is a first type electricity user and the characteristic electricity consumption period is a second type electricity consumption period, the power company increases the electricity transaction price of the regional electricity user during the characteristic electricity consumption period; When the regional electricity user is a second type of electricity user, the power company normally conducts electricity transactions with the regional electricity user.

10. An electricity trading system based on multi-party secure computing and federated learning, characterized in that: include: The data acquisition module is used to acquire the power trading area and respectively acquire the power user historical monitoring data corresponding to each regional power user in the power trading area; The historical monitoring data of the power consumption unit includes: abnormal power cycle number ratio, cycle average power consumption deviation and cycle power consumption deviation variance; A unit analysis module, used to classify regional power users in the power trading area into first type power users and second type power users according to historical monitoring data of power users; The time period analysis module is used to mark a power consumption characteristic time period in the future power consumption time period of the first type of power consumption unit according to the power consumption unit type classification result, and obtain the time period power consumption analysis data by performing power analysis on the power consumption characteristic time period to obtain the time period estimated power consumption, price sensitivity evaluation coefficient and power supply index value corresponding to each first type of power consumption unit in the power consumption characteristic time period; The power trading module is used to divide multiple power consumption characteristic time periods into first type power consumption time periods and second type power consumption time periods according to the power consumption analysis data of the time periods, and to control power trading of regional power consumption units according to the power consumption unit type division results and the power consumption characteristic time period division results.

Images