Electricity utilization abnormity monitoring system for electricity utilization user

Abstract

The invention, which relates to the field of power monitoring, discloses a power utilization abnormity monitoring system for a power user, comprising a control center which is connected with a data acquisition module, a personal power model establishment module, a model sharing pool module, a data processing module and an abnormity early warning module. The data acquisition module is used for acquiring power data of a power user; the personal power model establishing module is used for establishing a personal power model of the power user according to the power data; the model sharing pool module is used for generating a model sharing pool according to the personal power model; the data processing module is used for analyzing the power data of the power users according to the model sharing pool and the personal power model, and generating corresponding abnormal early warning according to the analysis result; the abnormity early warning module is used for processing abnormity early warning.

Description

Technical Field

[0001] The invention relates to the technical field of power monitoring, and in particular to a power consumption abnormality monitoring system for power users. Background Art

[0002] With the continuous development of society and the advancement of intelligent technology, the importance of power systems in daily life and industrial production has become increasingly prominent. Traditional power monitoring and management models usually rely on manual inspections and basic power data collection methods, and there are problems such as delayed data analysis, inadequate abnormal monitoring, and slow response speed. Especially in the case of large-scale user distribution and changing electricity demand, how to achieve efficient, accurate and intelligent management of the power system has become a technical problem that needs to be solved urgently.

[0003] At present, although some electricity monitoring and early warning systems have been applied in the market, most systems still rely on traditional single models or basic anomaly detection methods, which are difficult to cope with the complex electricity demand and personalized scenarios of different users. How to accurately capture the changes in electricity users' electricity consumption habits and improve the sensitivity and accuracy of anomaly detection is a problem we need to consider at present. Summary of the invention

[0004] In order to solve the above problems, an object of the present invention is to provide a system for monitoring abnormal power consumption of power users.

[0005] The object of the present invention can be achieved by the following technical solutions: A power consumption abnormality monitoring system for power users, comprising a control center, wherein the control center is communicatively connected with a data acquisition module, a personal power model building module, a model sharing pool module, a data processing module and an abnormality warning module;

[0006] The data collection module is used to collect power data of electricity users;

[0007] The personal power model building module is used to create a personal power model of the power user according to the power data;

[0008] The model sharing pool module is used to generate a model sharing pool according to the personal power model;

[0009] The data processing module is used to analyze the power data of the power user according to the model sharing pool and the personal power model, and generate corresponding abnormal warning according to the analysis results;

[0010] The abnormal warning module is used to process abnormal warnings.

[0011] Furthermore, the process of the data collection module collecting power data of power users includes:

[0012] The power data includes power consumption and power consumption time;

[0013] Set up an electricity meter to obtain the electricity consumption of electricity - using users through the electricity meter;

[0014] Set a peak - period determination cycle;

[0015] Obtain the electricity consumption in real - time within the peak - period determination cycle, and mark the time when the electricity consumption is obtained as the electricity - using time, that is, there is a one - to - one correspondence between electricity consumption and electricity - using time;

[0016] Generate electricity - using coordinates based on electricity consumption and electricity - using time, where the electricity - using coordinate=(electricity - using time, electricity consumption);

[0017] Establish a two - dimensional coordinate system A, map the electricity - using coordinates into the two - dimensional coordinate system A, and connect the electricity - using coordinates with adjacent electricity - using times in the two - dimensional coordinate system A in sequence to generate several electricity - using line segments;

[0018] Set up electricity - change interval A, electricity - change interval B, and electricity - change interval C;

[0019] Obtain the slope of the electricity - using line segment;

[0020] Generate an electricity - using peak period based on the electricity - using time corresponding to the electricity - using line segment whose slope ∈ electricity - change interval A;

[0021] Generate an electricity - using flat - peak period based on the electricity - using time corresponding to the electricity - using line segment whose slope ∈ electricity - change interval B;

[0022] Generate an electricity - using valley - peak period based on the electricity - using time corresponding to the electricity - using line segment whose slope ∈ electricity - change interval C;

[0023] Set up collection period A, collection period B, collection period C, and a personal power database;

[0024] When the time after the peak - period determination cycle is in the electricity - using peak period, obtain the user's power data according to collection period A; when the time after the peak - period determination cycle is in the electricity - using flat - peak period, obtain the user's power data according to collection period B; when the time after the peak - period determination cycle is in the electricity - using valley - peak period, obtain the user's power data according to collection period C;

[0025] Store the power data into the personal power database.

[0026] Furthermore, the process by which the personal power model establishment module creates a personal power model for an electricity - using user based on power data includes:

[0027] Set up a construction period;

[0028] Regularly create a personal power model through the personal power database according to the construction period;

[0029] Regularly obtain the power data in the personal power database according to the construction period;

[0030] Set a period threshold. If there is a time interval between adjacent power consumption times in the power data that is greater than the period threshold, interpolation processing is performed within the time interval; otherwise, no operation is required.

[0031] That is, there are several power consumption times in the power data, power consumption time 1, power consumption time 2,..., power consumption time n , where n is a positive integer; if there is a power consumption time i and power consumption time i+1 and the time interval between them is greater than the period threshold, then interpolation processing is performed within the time interval between power consumption time i and power consumption time i+1 ; otherwise, no operation is required. Among them, the time interval = power consumption time i+1 - power consumption time i ;

[0032] Set the regular power consumption period.

[0033] Perform data denoising and data smoothing processing on the power data to generate new power data.

[0034] Establish a two-dimensional coordinate system B, generate new power consumption coordinates according to the new power data, map the new power consumption coordinates to the two-dimensional coordinate system B, and generate new power consumption line segments according to the power consumption times in the new power consumption coordinates.

[0035] Obtain the power consumption time corresponding to the regular power consumption period, and then obtain the power consumption corresponding to the power consumption time according to the new power consumption line segment. Generate regular power consumption coordinates according to the power consumption time corresponding to the regular power consumption period and its corresponding power consumption in the new power consumption line segment.

[0036] The regular power consumption coordinates = (power consumption time, power consumption), where the power consumption time is the power consumption time corresponding to the regular power consumption period, and the power consumption is the power consumption corresponding to the power consumption time in the new power consumption line segment.

[0037] Generate a personal power model according to the regular power consumption coordinates through machine learning.

[0038] Furthermore, the process of interpolation processing is as follows:

[0039] Assume that there is a time interval between adjacent power consumption times i and power consumption time i+1 in the power data that is greater than the period threshold, where i is a positive integer;

[0040] Mark the power consumption corresponding to power consumption time i and power consumption time i+1 as power consumption i and power consumptioni+1 ;

[0041] Generate interpolation values based on the power consumption time i , the power consumption time i+1 and the cycle threshold;

[0042]

[0043] Divide the time interval between the power consumption time i and the power consumption time i+1 equally according to the interpolation values, and generate a number of interpolated power consumption times based on the equal division results;

[0044] Generate a number of interpolated power consumptions based on the power consumption i , the power consumption i+1 and the mean difference value;

[0045] Obtain the interpolated power consumption time and its corresponding interpolated power consumption, and generate interpolated power consumption data; the mean difference power consumption data includes the power consumption time and the power consumption, where the power consumption time is the interpolated power consumption time, the power consumption is the interpolated power consumption, and the power consumption time and the power consumption are in a one-to-one correspondence, that is, the interpolated power consumption time and the interpolated power consumption are in a one-to-one correspondence;

[0046] Insert the interpolated power consumption data into the power data according to the power consumption time, and complete the interpolation processing within the time interval between the power consumption time i and the power consumption time i+1 ;

[0047] where t is a positive integer, and the power consumption time i < the interpolated power consumption time t < the power consumption time i+1 ;

[0048] where t is a positive integer, and the power consumption i < the interpolated power consumption t < the power consumption i+1 ;

[0049] The interpolated power consumption time t and the interpolated power consumption t are in a one-to-one correspondence, that is, the interpolated power consumption time 1 corresponds to the interpolated power consumption 1, the interpolated power consumption time 2 corresponds to the interpolated power consumption 2,..., the interpolated power consumption time n corresponds to the interpolated power consumption n correspondingly, where n is a positive integer;

[0050] Insert the interpolated power consumption data into the power data according to the power consumption time, that is, insert the interpolated power consumption time t into the power consumption time iand electricity consumption time i+1 Among them, the average inserted electricity consumption t is used to average the electricity consumption time t Insert the electricity consumption time i and the electricity consumption time i+1 Insert the electricity consumption amount in sequence between them i and the electricity consumption amount i+1 in between.

[0051] Furthermore, the process by which the model sharing pool module generates the model sharing pool according to the individual power model includes:

[0052] Generate an individual power code according to the electricity user, the individual power code is unique, and the individual power model and the individual power code are in a corresponding relationship;

[0053] The model sharing pool includes a model upload pool and a model framework pool;

[0054] Regularly obtain the individual power models of each electricity user according to the construction period, and submit the individual power models to the model upload pool;

[0055] Set the model refinement period;

[0056] Regularly generate a power framework model according to the individual power models in the model upload pool according to the model refinement period, and then submit the power framework model to the model framework pool.

[0057] Furthermore, the process of generating a power framework model according to the individual power models in the model upload pool is as follows:

[0058] Establish a two-dimensional coordinate system C;

[0059] Obtain the corresponding regular electricity consumption coordinates of the individual power models in the model upload pool, map the regular electricity consumption coordinates to the two-dimensional coordinate system C, and generate regular electricity consumption line segments according to the electricity consumption time in the regular electricity consumption coordinates;

[0060] Obtain the slope of the regular electricity consumption line segment according to the regular electricity consumption coordinates, and generate a slope change rate according to the slope;

[0061] where i is a positive integer;

[0062] If there are several regular electricity consumption line segments, regular electricity consumption line segment 1, regular electricity consumption line segment 2,..., regular electricity consumption line segment n , where n is a positive integer; obtain the slopes of the regular electricity consumption line segments according to the regular electricity consumption coordinates, then there are several slopes, slope 1, slope 2,..., slope n , where n is a positive integer, and slope 1 is the slope corresponding to regular electricity consumption line segment 1, slope 2 is the slope corresponding to regular electricity consumption line segment 2,..., slope nIs a conventional power consumption line segment n The corresponding slope; generate a ramp rate according to the slope, where i is a positive integer;

[0063] Obtain the ramp timing and ramp property corresponding to the ramp rate;

[0064] The ramp rate, ramp timing, and ramp property are in one-to-one correspondence;

[0065] The said ramp timing i / i+1 = [power consumption time a , power consumption time b , where the power consumption time a is the minimum power consumption time corresponding to the conventional power consumption coordinates of the slope that generates this ramp rate, and the power consumption time b is the maximum power consumption time corresponding to the conventional power consumption coordinates of the slope that generates this ramp rate;

[0066] The said ramp property i / i+1 includes "+", "-", "+-", and "0". If the slope i / i+1 corresponding to the ramp rate i and the slope i+1 are both greater than 0, then the ramp property i / i+1 corresponding to this ramp rate is "+", if the slope i / i+1 corresponding to the ramp rate i and the slope i+1 are both 0, then the ramp property i / i+1 corresponding to this ramp rate is "0", if the slope i / i+1 corresponding to the ramp rate i and the slope i+1 are both less than 0, then the ramp property i / i+1 corresponding to this ramp rate is "-", otherwise, the ramp property i / i+1 corresponding to this ramp rate is "+-";

[0067] Set the slope difference interval and the model general threshold;

[0068] There exist personal power models P and Q;

[0069] The ramp rate P corresponding to the personal power model P is ramp rate P 1 / 2 , ramp rate P 2 / 3 , ……, ramp rate P n-1 / n , the ramp timing P is ramp timing P 1 / 2 , ramp timing P 2 / 3 , ……, ramp timing P n-1 / n , and the ramp property P is ramp property P 1 / 2 , ramp property P 2 / 3 , ……, ramp property Pn-1 / n , where n is a positive integer;

[0070] The slope change rate Q corresponding to the personal power model Q is the slope change rate Q 1 / 2 , slope change rate Q 2 / 3 , ……, slope change rate Q n-1 / n , the slope change timing Q is the slope change timing Q 1 / 2 , slope change timing Q 2 / 3 , ……, slope change timing Q n-1 / n , the slope change property Q is the slope change property Q 1 / 2 , slope change property Q 2 / 3 , ……, slope change property Q n-1 / n , where n is a positive integer;

[0071] If there are consecutive num identical slope change timings in the personal power model P and the personal power model Q, and the corresponding slope change properties are the same and the differences in slope change rates are all within the slope difference interval, then electronic features are generated based on these consecutive num slope change timings and their corresponding slope change properties and slope change rates in the personal power model P and the personal power model Q respectively; where num is greater than or equal to the model general threshold;

[0072] The difference in slope change rate = |slope change rate P i / i+1 - slope change rate Q i / i+1 |, where i is a positive integer;

[0073] Then there are consecutive num slope change timings in the personal power model P and the personal power model Q, and the corresponding slope change properties are the same and the differences in slope change rates are all within the slope difference interval, that is, there are slope change timings in the personal power model P and the personal power model Q i+1 / i+2 , slope change timing i+2 / i+3 , …… and slope change timing i+num-1 / i+num , slope change property P i+1 / i+2 and slope change property Q i+1 / i+2 are equal, slope change property P i+2 / i+3 and slope change property Q i+2 / i+3 are equal, ……, slope change property P i+num-1 / i+num and slope change property Q i+num-1 / i+num are equal, and, the difference in slope change rate corresponding to the slope change timing i+1 / i+2 is within the slope difference interval, the difference in slope change rate corresponding to the slope change timing i+2 / i+3 is within the slope difference interval, ……, the difference in slope change rate corresponding to the slope change timing i+num-1 / i+num is within the slope difference interval;

[0074] Set the feature frame threshold A and the feature frame threshold B;

[0075] Obtain the slew rate, slew time series, and slew property corresponding to each individual power model in the model upload pool, generate electronic features based on the slew rate, slew time series, and slew property, and then generate a power framework model based on each individual power model and its corresponding electronic features;

[0076] If num2 identical electronic features can be generated based on the slew rate, slew time series, and slew property corresponding to num1 individual power models, then a power framework model is generated through machine learning based on the electronic features; where num1 ≥ feature framework threshold A and num2 ≥ feature framework threshold B.

[0077] Further, the process of the data processing module analyzing the power data of electricity users based on the model sharing pool and individual power models and generating corresponding anomaly warnings includes:

[0078] Obtain the power data of electricity users in real time;

[0079] Obtain the individual power code of the electricity user, and obtain the corresponding individual power model and its corresponding power framework model in the model sharing pool according to the individual power code;

[0080] If electricity user A and electricity user B are on the same distribution line, then electricity user A and electricity user B are local users to each other; denote the individual power codes of local users as local power codes;

[0081] Judge whether the power data obtained in real time is abnormal according to the individual power model. If there is no abnormality, no operation is required. If there is an abnormality, judge whether the power data obtained in real time is abnormal according to the power framework model. If there is no abnormality, no operation is required. If there is an abnormality, obtain the individual power model corresponding to the local power code and the real-time power data of the local user, and judge whether the real-time power data of the local user is abnormal according to the individual power model corresponding to the local power code. If all are abnormal, generate a first-level anomaly warning and send the first-level anomaly warning to the warning module. Otherwise, generate a second-level anomaly warning and send the second-level anomaly warning to the warning module.

[0082] Further, the process of the anomaly warning module for processing anomaly warnings includes:

[0083] If a first-level anomaly warning is received, inform the electricity user that there may be an abnormality in the distribution line and let the electricity user make corresponding handling;

[0084] If a second-level anomaly warning is received, inform the electricity user that there is an abnormal power consumption and let the electricity user make corresponding handling.

[0085] Compared with the prior art, the beneficial effects of the present invention are:

[0086] 1. The present invention sets different acquisition cycles according to different peak periods of electricity consumption of electricity users, saving system resources while ensuring the accuracy of the collected power data;

[0087] 2. The present invention regularly creates personal power models of electricity users according to the construction cycle, ensuring the timeliness of the personal power models and improving the accuracy of subsequent judgment on whether there are abnormalities in power data based on the personal power models;

[0088] 3. The present invention generates a power framework model based on the personal power model, and judges whether there are abnormalities in power data through the power framework model, improving the fault tolerance rate of the model and reducing the possibility of misjudgment caused by normal fluctuations of power data;

[0089] 4. The present invention reduces the possibility of misjudging the electricity consumption abnormality of electricity users by power distribution cables by setting the possibility of local users checking whether there are abnormalities in the power distribution lines. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] Figure 1 is the schematic diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0091] As Figure 1 shown, an electricity consumption abnormality monitoring system for electricity users includes a control center, and the control center is connected with a data acquisition module, a personal power model establishment module, a model sharing pool module, a data processing module and an abnormality warning module;

[0092] The data acquisition module is used to acquire the power data of electricity users;

[0093] The personal power model establishment module is used to create a personal power model of an electricity user according to the power data;

[0094] The model sharing pool module is used to generate a model sharing pool according to the personal power model;

[0095] The data processing module is used to analyze the power data of electricity users according to the model sharing pool and the personal power model, and generate corresponding abnormality warnings according to the analysis results;

[0096] The abnormality warning module is used to process the abnormality warnings;

[0097] It should be further noted that, in the specific implementation process, the process of the data acquisition module acquiring the power data of electricity users includes:

[0098] The power data includes power consumption and power consumption time;

[0099] An electric energy meter is set, and the power consumption of the electricity user is obtained through the electric energy meter;

[0100] Set the peak determination period;

[0101] Obtain the electricity consumption within the peak determination period in real time, and mark the time when the electricity consumption is obtained as the power consumption time, that is, the electricity consumption and the power consumption time are in a one-to-one correspondence relationship;

[0102] Generate power consumption coordinates based on the electricity consumption and the power consumption time, where the power consumption coordinates = (power consumption time, electricity consumption);

[0103] Establish a two-dimensional coordinate system A, map the power consumption coordinates into the two-dimensional coordinate system A, and connect the power consumption coordinates with adjacent power consumption times in the two-dimensional coordinate system A in sequence according to the power consumption time to generate a number of power consumption line segments;

[0104] Set the electricity change interval A, the electricity change interval B, and the electricity change interval C;

[0105] Obtain the slope of the power consumption line segment;

[0106] Generate the electricity peak period according to the power consumption time corresponding to the power consumption line segment with the slope ∈ the electricity change interval A;

[0107] Generate the electricity flat peak period according to the power consumption time corresponding to the power consumption line segment with the slope ∈ the electricity change interval B;

[0108] Generate the electricity valley peak period according to the power consumption time corresponding to the power consumption line segment with the slope ∈ the electricity change interval C;

[0109] Set the collection period A, the collection period B, the collection period C, and the personal power database;

[0110] When the time after the peak determination period is in the electricity peak period, obtain the user's power data according to the collection period A; when the time after the peak determination period is in the electricity flat peak period, obtain the user's power data according to the collection period B; when the time after the peak determination period is in the electricity valley peak period, obtain the user's power data according to the collection period C;

[0111] Store the power data in the personal power database;

[0112] It should be further noted that in the specific implementation process, the process of the personal power model establishment module creating the personal power model of the power consumption user includes:

[0113] Set the construction period;

[0114] Create a personal power model regularly through the personal power database according to the construction period;

[0115] Regularly obtain the power data in the personal power database according to the construction period;

[0116] Set a periodic threshold. If the time interval between adjacent power consumption times in the power data is greater than the periodic threshold, interpolation processing is performed within the time interval; otherwise, no operation is required.

[0117] That is, there are several power consumption times in the power data, power consumption time 1, power consumption time 2,..., power consumption time n , where n is a positive integer; if there exists a power consumption time i and power consumption time i+1 and the time interval between them is greater than the periodic threshold, then interpolation processing is performed within the time interval between power consumption time i and power consumption time i+1 ; otherwise, no operation is required. Among them, the time interval = power consumption time i+1 - power consumption time i ;

[0118] Set the regular power consumption period.

[0119] Perform data denoising and data smoothing on the power data to generate new power data.

[0120] Establish a two-dimensional coordinate system B, generate new power consumption coordinates according to the new power data, map the new power consumption coordinates to the two-dimensional coordinate system B, and generate new power consumption line segments according to the power consumption times in the new power consumption coordinates.

[0121] Obtain the power consumption time corresponding to the regular power consumption period, and then obtain the power consumption corresponding to the power consumption time according to the new power consumption line segment. Generate regular power consumption coordinates according to the power consumption time corresponding to the regular power consumption period and its corresponding power consumption in the new power consumption line segment.

[0122] The regular power consumption coordinate = (power consumption time, power consumption), where the power consumption time is the power consumption time corresponding to the regular power consumption period, and the power consumption is the power consumption corresponding to the power consumption time in the new power consumption line segment.

[0123] Generate a personal power model based on the regular power consumption coordinates through machine learning.

[0124] Among them, the process of interpolation processing is as follows:

[0125] Assume that there is a time interval greater than the periodic threshold between adjacent power consumption times i and power consumption time i+1 in the power data, where i is a positive integer;

[0126] Mark the power consumptions corresponding to power consumption time i and power consumption time i+1 as power consumption i and power consumption i+1 respectively;

[0127] Generate interpolation values based on the power consumption time i and the power consumption time i+1 and the cycle threshold;

[0128]

[0129] Divide the time interval between the power consumption time i and the power consumption time i+1 equally according to the interpolation values, and generate a number of interpolated power consumption times based on the equal division results;

[0130] Generate a number of interpolated power consumptions according to the power consumption i and the power consumption i+1 and the mean difference value;

[0131] Obtain the interpolated power consumption time and its corresponding interpolated power consumption, and generate interpolated power consumption data; the mean difference power consumption data includes the power consumption time and the power consumption, where the power consumption time is the interpolated power consumption time, the power consumption is the interpolated power consumption, and the power consumption time and the power consumption are in a one-to-one correspondence, that is, the interpolated power consumption time and the interpolated power consumption are in a one-to-one correspondence;

[0132] Insert the interpolated power consumption data into the power data according to the power consumption time, and complete the interpolation process within the time interval between the power consumption time i and the power consumption time i+1 ;

[0133] where t is a positive integer, and the power consumption time i < the interpolated power consumption time t < the power consumption time i+1 ;

[0134] where t is a positive integer, and the power consumption i < the interpolated power consumption t < the power consumption i+1 ;

[0135] The interpolated power consumption time t and the interpolated power consumption t are in a one-to-one correspondence, that is, the interpolated power consumption time 1 corresponds to the interpolated power consumption 1, the interpolated power consumption time 2 corresponds to the interpolated power consumption 2,..., the interpolated power consumption time n corresponds to the interpolated power consumption n where n is a positive integer;

[0136] Insert the interpolated power consumption data into the power data according to the power consumption time, that is, insert the interpolated power consumption time t into the power consumption time i and the power consumption time i+1 in sequence according to the power consumption time, and insert the interpolated power consumptiont Equalize the electricity consumption time t Insert the electricity consumption time i And the electricity consumption time i+1 Insert the electricity consumption amount in sequence between the electricity consumption time i And the electricity consumption amount i+1 Between them;

[0137] It should be further noted that in the specific implementation process, the process of the model sharing pool module generating the model sharing pool according to the personal power model includes:

[0138] Generate a personal power code according to the electricity user, the personal power code is unique, and the personal power model and the personal power code are in a corresponding relationship;

[0139] The model sharing pool includes a model upload pool and a model framework pool;

[0140] Regularly obtain the personal power models of each electricity user according to the construction period, and submit the personal power models to the model upload pool;

[0141] Set the model refinement period;

[0142] Regularly generate a power framework model according to the personal power models in the model upload pool according to the model refinement period, and then submit the power framework model to the model framework pool;

[0143] Among them, the process of generating a power framework model according to the personal power models in the model upload pool is:

[0144] Establish a two-dimensional coordinate system C;

[0145] Obtain the corresponding regular electricity consumption coordinates of the personal power models in the model upload pool, map the regular electricity consumption coordinates to the two-dimensional coordinate system C, and generate regular electricity consumption line segments according to the electricity consumption time in the regular electricity consumption coordinates;

[0146] Obtain the slope of the regular electricity consumption line segment according to the regular electricity consumption coordinates, and generate a slope change rate according to the slope;

[0147] Among them, i is a positive integer;

[0148] If there are several regular electricity consumption line segments, regular electricity consumption line segment 1, regular electricity consumption line segment 2,..., regular electricity consumption line segment n , where n is a positive integer; obtain the slope of the regular electricity consumption line segment according to the regular electricity consumption coordinates, then there are several slopes, slope 1, slope 2,..., slope n , where n is a positive integer, and slope 1 is the slope corresponding to regular electricity consumption line segment 1, slope 2 is the slope corresponding to regular electricity consumption line segment 2,..., slope n Is the regular electricity consumption line segment nThe corresponding slope; generating a ramp rate according to the slope, where i is a positive integer;

[0149] Obtaining the ramp timing and ramp property corresponding to the ramp rate;

[0150] The ramp rate, ramp timing, and ramp property are in a one-to-one correspondence;

[0151] The said ramp timing i / i+1 = [power consumption time a , power consumption time b , where the power consumption time a is the minimum power consumption time among the power consumption times corresponding to the normal power consumption coordinates of the slope corresponding to this ramp rate, and the power consumption time b is the maximum power consumption time among the power consumption times corresponding to the normal power consumption coordinates of the slope corresponding to this ramp rate;

[0152] The said ramp property i / i+1 includes "+", "-", "+-", and "0". If the slopes i / i+1 corresponding to the ramp rate i and the slope i+1 are both greater than 0, then the ramp property i / i+1 corresponding to this ramp rate is "+", if the slopes i / i+1 corresponding to the ramp rate i and the slope i+1 are both 0, then the ramp property i / i+1 corresponding to this ramp rate is "0", if the slopes i / i+1 corresponding to the ramp rate i and the slope i+1 are both less than 0, then the ramp property i / i+1 corresponding to this ramp rate is "-", otherwise, the ramp property i / i+1 corresponding to this ramp rate is "+-";

[0153] Setting a slope difference interval and a model general threshold;

[0154] There exist a personal power model P and a personal power model Q;

[0155] The ramp rates P corresponding to the personal power model P are ramp rate P 1 / 2 , ramp rate P 2 / 3 , ……, ramp rate P n-1 / n , the ramp timings P are ramp timing P 1 / 2 , ramp timing P 2 / 3 , ……, ramp timing P n-1 / n , and the ramp properties P are ramp property P 1 / 2 , ramp property P 2 / 3 , ……, ramp property P n-1 / n , where n is a positive integer;

[0156] The slope change rate Q corresponding to the personal power model Q is the slope change rate Q 1 / 2 and the slope change rate Q 2 / 3 and so on, and the slope change rate Q n-1 / n The slope change timing Q is the slope change timing Q 1 / 2 and the slope change timing Q 2 / 3 and so on, and the slope change timing Q n-1 / n The slope change property Q is the slope change property Q 1 / 2 and the slope change property Q 2 / 3 and so on, and the slope change property Q n-1 / n where n is a positive integer;

[0157] If there are consecutive num identical slope change timings in the personal power model P and the personal power model Q, and the corresponding slope change properties are the same and the differences in slope change rates are all within the slope difference interval, then generate electronic features based on these consecutive num slope change timings and their corresponding slope change properties and slope change rates in the personal power model P and the personal power model Q respectively; where num is greater than or equal to the model general threshold;

[0158] The difference in slope change rates = |slope change rate P i / i+1 - slope change rate Q i / i+1 |, where i is a positive integer;

[0159] Then, if there are consecutive num slope change timings in the personal power model P and the personal power model Q, and the corresponding slope change properties are the same and the differences in slope change rates are all within the slope difference interval, that is, there are slope change timings i+1 / i+2 and slope change timings i+2 / i+3 and so on, and slope change timings i+num-1 / i+num in the personal power model P and the personal power model Q, the slope change property P i+1 / i+2 is equal to the slope change property Q i+1 / i+2 , the slope change property P i+2 / i+3 is equal to the slope change property Q i+2 / i+3 and so on, and the slope change property P i+num-1 / i+num is equal to the slope change property Q i+num-1 / i+num . And, the difference in slope change rates corresponding to the slope change timing i+1 / i+2 is within the slope difference interval, the difference in slope change rates corresponding to the slope change timing i+2 / i+3 is within the slope difference interval, and so on, and the difference in slope change rates corresponding to the slope change timing i+num-1 / i+num is within the slope difference interval;

[0160] Set the feature frame threshold A and the feature frame threshold B;

[0161] Obtain the slew rate, slew time series, and slew properties corresponding to each individual power model in the model upload pool, generate electronic features based on the slew rate, slew time series, and slew properties, and then generate a power framework model based on each individual power model and its corresponding electronic features;

[0162] If num2 identical electronic features can be generated based on the slew rate, slew time series, and slew properties corresponding to num1 individual power models, then generate a power framework model based on the electronic features through machine learning; where num1 ≥ feature framework threshold A and num2 ≥ feature framework threshold B;

[0163] It should be further noted that in the specific implementation process, the process of the data processing module analyzing the power data of electricity users based on the model sharing pool and individual power models and generating corresponding abnormal warnings includes:

[0164] Obtain the power data of electricity users in real time;

[0165] Obtain the individual power code of the electricity user, and obtain the corresponding individual power model and its corresponding power framework model in the model sharing pool according to the individual power code;

[0166] If electricity user A and electricity user B are on the same distribution line, then electricity user A and electricity user B are local users to each other; record the individual power code of the local user as the local power code;

[0167] Judge whether the power data obtained in real time is abnormal according to the individual power model. If there is no abnormality, no operation is required. If there is an abnormality, judge whether the power data obtained in real time is abnormal according to the power framework model. If there is no abnormality, no operation is required. If there is an abnormality, obtain the individual power model corresponding to the local power code and the real-time power data of the local user, and judge whether the real-time power data of the local user is abnormal according to the individual power model corresponding to the local power code. If all are abnormal, generate a first-level abnormal warning and send the first-level abnormal warning to the warning module. Otherwise, generate a second-level abnormal warning and send the second-level abnormal warning to the warning module;

[0168] It should be further noted that in the specific implementation process, the process of the abnormal warning module for processing abnormal warnings includes:

[0169] If a first-level abnormal warning is received, inform the electricity user that there may be an abnormality in the distribution line and let the electricity user make corresponding handling;

[0170] If a second-level abnormal warning is received, inform the electricity user that there is an abnormal power consumption and let the electricity user make corresponding handling;

[0171] The above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An abnormal power consumption monitoring system for electricity users, including a control center, characterized in that, The control center is communicatively connected to a data acquisition module, a personal power model establishment module, a model sharing pool module, a data processing module, and an anomaly warning module; The data acquisition module is used to acquire the power data of electricity-using users; The personal power model establishment module is used to create a personal power model of electricity-using users according to the power data; The model sharing pool module is used to generate a model sharing pool according to the personal power model; The data processing module is used to analyze the power data of electricity-using users according to the model sharing pool and the personal power model, and generate corresponding anomaly warnings according to the analysis results; The anomaly warning module is used to process the anomaly warnings.

2. The power consumption anomaly monitoring system for electricity users according to claim 1, wherein, The process by which the data acquisition module acquires the power data of electricity-using users includes: The power data includes power consumption and power consumption time; An electricity meter is set up to obtain the power consumption of electricity-using users through the electricity meter; A peak period determination period is set; The power consumption within the peak period determination period is obtained in real time, and the time when the power consumption is obtained is marked as the power consumption time, that is, the power consumption and the power consumption time have a one-to-one correspondence; An electricity consumption coordinate is generated according to the power consumption and the power consumption time, and the electricity consumption coordinate = (power consumption time, power consumption); A two-dimensional coordinate system A is established, the electricity consumption coordinates are mapped into the two-dimensional coordinate system A, and the electricity consumption coordinates adjacent in terms of power consumption time in the two-dimensional coordinate system A are connected pairwise in sequence to generate a number of electricity consumption line segments; An electricity change interval A, an electricity change interval B, and an electricity change interval C are set; The slope of the electricity consumption line segment is obtained; An electricity consumption peak period is generated according to the power consumption time corresponding to the electricity consumption line segment where the slope ∈ electricity change interval A; An electricity consumption flat peak period is generated according to the power consumption time corresponding to the electricity consumption line segment where the slope ∈ electricity change interval B; An electricity consumption valley peak period is generated according to the power consumption time corresponding to the electricity consumption line segment where the slope ∈ electricity change interval C; A collection period A, a collection period B, a collection period C, and a personal power database are set; When the time after the peak period determination period is in the electricity consumption peak period, the power data of the user is obtained according to the collection period A; when the time after the peak period determination period is in the electricity consumption flat peak period, the power data of the user is obtained according to the collection period B; when the time after the peak period determination period is in the electricity consumption valley peak period, the power data of the user is obtained according to the collection period C; The power data is stored in the personal power database.

3. The power consumption anomaly monitoring system for electricity users according to claim 2, characterized in that, The process by which the personal power model establishment module creates a personal power model of electricity-using users according to the power data includes: A construction period is set; The power data in the personal power database is obtained regularly according to the construction period; A period threshold is set. If there is a time interval between adjacent power consumption times in the power data that is greater than the period threshold, interpolation processing is performed within the time interval, otherwise, no operation is required; An electricity consumption regular period is set; The power data is subjected to data denoising processing and data smoothing processing to generate new power data; A two-dimensional coordinate system B is established, new electricity consumption coordinates are generated according to the new power data, the new electricity consumption coordinates are mapped into the two-dimensional coordinate system B, and new electricity consumption line segments are generated according to the power consumption time in the new electricity consumption coordinates; Obtain the electricity consumption time corresponding to the regular electricity consumption cycle, and then obtain the electricity consumption corresponding to the electricity consumption time according to the new electricity consumption segment. Generate regular electricity consumption coordinates based on the electricity consumption time corresponding to the regular electricity consumption cycle and the electricity consumption corresponding to it in the new electricity consumption segment; the regular electricity consumption coordinates = (electricity consumption time, electricity consumption), where the electricity consumption time is the electricity consumption time corresponding to the regular electricity consumption cycle, and the electricity consumption is the electricity consumption corresponding to the electricity consumption time in the new electricity consumption segment; Generate a personal power model through machine learning based on the regular electricity consumption coordinates.

4. An abnormal power consumption monitoring system for power users according to claim 3, characterized in that, The process of interpolation processing is as follows: Assume that there is an adjacent power consumption time in the power data i and the power consumption time i+1 with a time interval greater than the period threshold, where i is a positive integer; mark the power consumptions corresponding to the power consumption time i and the power consumption time i+1 as the power consumptions i and the power consumption i+1 respectively; According to the power consumption time i and the power consumption time i+1 generate the average interpolation value based on the cycle threshold; Divide the time interval between the electricity consumption time i and the electricity consumption time i+1 equally according to the evenly interpolated values, and generate a number of evenly interpolated electricity consumption times based on the equal division results; According to the power consumption i and the power consumption i+1 and the mean difference value, generate a number of interpolated power consumptions; Obtain the evenly interpolated electricity consumption time and its corresponding evenly interpolated electricity consumption, and generate evenly interpolated electricity consumption data; Insert the evenly inserted power consumption data into the power data according to the power consumption time to complete the power consumption time i and the power consumption time i+1 Perform interpolation processing for the time interval between them.

5. The power consumption anomaly monitoring system for electricity users according to claim 4, characterized in that The process by which the model sharing pool module generates the model sharing pool based on the personal power model includes: Generate a personal power code according to the electricity consumption user. The personal power code is unique, and the personal power model and the personal power code are in a corresponding relationship; The model sharing pool includes a model upload pool and a model framework pool; Regularly obtain the personal power models of each electricity consumption user according to the construction period, and submit the personal power models to the model upload pool; Set the model refinement period; regularly generate a power framework model according to the personal power models in the model upload pool according to the model refinement period, and then submit the power framework model to the model framework pool.

6. The power consumption anomaly monitoring system for electricity users according to claim 5, characterized in that, The process of generating a power framework model according to the personal power models in the model upload pool is as follows: Establish a two-dimensional coordinate system C; Obtain the regular electricity consumption coordinates corresponding to the personal power models in the model upload pool, map the regular electricity consumption coordinates to the two-dimensional coordinate system C, and generate a regular electricity consumption segment according to the electricity consumption time in the regular electricity consumption coordinates; Obtain the slope of the regular electricity consumption segment according to the regular electricity consumption coordinates, and generate a slope change rate according to the slope; The said Obtain the slope change time sequence and slope change property corresponding to the slope change rate; The slope change rate, slope change time sequence, and slope change property are in a one-to-one correspondence relationship; The ramped time series i / i+1 = [power consumption time a , power consumption time b , where the power consumption time a is the minimum power consumption time corresponding to the conventional power consumption coordinates for the slope that generates this ramp rate, and the power consumption time b is the maximum power consumption time corresponding to the conventional power consumption coordinates for the slope that generates this ramp rate; The ramp property i / i+1 including "+", "-", "+-", and "0"; Set a slope difference interval and a model general threshold; There are a personal power model P and a personal power model Q; If there are consecutive num identical slope change time sequences in the personal power model P and the personal power model Q, and the corresponding slope change properties are the same and the differences in the slope change rates are all within the slope difference interval, then generate electronic features for electricity consumption according to the consecutive num slope change time sequences and the slope change properties and slope change rates corresponding to them in the personal power model P and the personal power model Q respectively; where num is greater than or equal to the model general threshold; The difference in ramp rates = |ramp rate P i / i+1 - ramp rate Q i / i+1 |, where i is a positive integer; Set a feature framework threshold A and a feature framework threshold B; Obtain the slope change rate, slope change time sequence, and slope change property corresponding to each personal power model in the model upload pool, generate electronic features for electricity consumption according to the slope change rate, slope change time sequence, and slope change property, and then generate a power framework model according to each personal power model and its corresponding electronic features for electricity consumption; If num2 identical electronic features for electricity consumption can be generated according to the slope change rate, slope change time sequence, and slope change property corresponding to num1 personal power models, then generate a power framework model according to the electronic features for electricity consumption through machine learning; where num1 ≥ feature framework threshold A and num2 ≥ feature framework threshold B.

7. An abnormal power consumption monitoring system for power users according to claim 6, characterized in that, The process in which the data processing module analyzes the power data of electricity users based on the model sharing pool and the personal power model and generates corresponding abnormal warnings includes: Obtaining the power data of electricity users in real time; Obtaining the personal power code of the electricity user, and obtaining the corresponding personal power model and its corresponding power framework model in the model sharing pool according to the personal power code; If electricity user A and electricity user B are on the same distribution line, then electricity user A and electricity user B are local users to each other; the personal power codes of the local users are recorded as local power codes; Judging whether the power data obtained in real time is abnormal according to the personal power model. If there is no abnormality, no operation is required. If there is an abnormality, judging whether the power data obtained in real time is abnormal according to the power framework model. If there is no abnormality, no operation is required. If there is an abnormality, obtaining the personal power model corresponding to the local power code and the real-time power data of the local user, and judging whether the real-time power data of the local user is abnormal according to the personal power model corresponding to the local power code. If all are abnormal, a first-level abnormal warning is generated, and then the first-level abnormal warning is sent to the warning module. Otherwise, a second-level abnormal warning is generated, and then the second-level abnormal warning is sent to the warning module.

8. The power consumption anomaly monitoring system for electricity users according to claim 7, characterized in that, The process in which the abnormal warning module is used to process abnormal warnings includes: If a first-level abnormal warning is received, it is informed to the electricity user that there may be an abnormality in the distribution line, and the electricity user makes corresponding handling; If a second-level abnormal warning is received, it is informed to the electricity user that the electricity consumption is abnormal, and the electricity user makes corresponding handling.

Images