Electricity consumption data anomaly analysis method and system

Abstract

The invention discloses a power consumption data anomaly analysis method and system. The method comprises the steps that whether the sequence deviation degree between a first target power consumption data sub-sequence and a second target power consumption data sub-sequence is larger than a preset deviation threshold value or not is judged; if the deviation value is greater than the preset deviation threshold value, respectively defining the deviation value as a first doubt power consumption data sub-sequence and a second doubt power consumption data sub-sequence, and performing power consumption data anomaly analysis on the first doubt power consumption data sub-sequence and the second doubt power consumption data sub-sequence according to a preset first data analysis strategy; according to the first doubt power consumption data sub-sequence and the second doubt power consumption data sub-sequence, and adopting a preset second data analysis strategy to perform abnormal power consumption data anomaly analysis on a first non-target power consumption data sub-sequence adjacent to the first doubt power consumption data sub-sequence and a second non-target power consumption data sub-sequence adjacent to the second doubt power consumption data sub-sequence. The abnormal power consumption data can be accurately determined.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power consumption data analysis, and particularly relates to a power consumption data anomaly analysis method and system. BACKGROUND

[0002] In the field of power equipment monitoring, accurate identification of abnormal power consumption data is crucial for preventing equipment failure, reducing fire risk, and ensuring power safety. The current mainstream anomaly detection methods mainly have the following technical bottlenecks: Existing technologies generally rely on single-dimensional power consumption threshold judgment (such as setting a fixed threshold or a statistical model based on historical data). However, during the device state switching process, normal load fluctuations (such as motor start-up, processing equipment feeding) and real anomalies (such as short circuit, insulation aging leading to rapid increase of current) will all show an increase in power consumption. Due to the lack of dynamic correlation analysis of real-time thermal state of the equipment, the existing methods have the following defects: High false positive rate: normal state switching (such as cold start current) is misreported as an anomaly; High risk of missed detection: when the abnormal current does not exceed the static threshold but has already caused rapid accumulation of device temperature (such as local poor contact), it cannot be timely warned.

[0003] Especially worth noting is that the change of device temperature has higher sensitivity to current anomaly - according to the Joule law (Q=I²Rt), the temperature rise rate caused by abnormal current is much higher than that of normal load change. However, the existing technology fails to effectively utilize the time series mutation characteristics of temperature data to accurately locate the suspicious interval of power consumption data, resulting in low analysis efficiency. SUMMARY

[0004] The application provides a power consumption data anomaly analysis method and system, which solves the technical problem that the time series mutation characteristics of temperature data are not effectively utilized to accurately locate the suspicious interval of power consumption data, resulting in low analysis efficiency.

[0005] In a first aspect, the application provides a power consumption data anomaly analysis method, comprising: obtaining power consumption data sequences and device temperature data sequences of power consumption equipment within a preset time period, wherein the power consumption data sequences contain power consumption data at different time nodes, and the device temperature data sequences contain device temperature data at different time nodes; selecting at least one mutation device temperature data in the device temperature data sequences according to the change trend of each device temperature data, and according to the time node of the at least one mutation device temperature data, obtaining at least one target power consumption data subsequence and at least one non-target power consumption data subsequence in the power consumption data sequences aligned with the device temperature data sequences based on a preset data extraction rule; determine whether a sequence deviation degree between the first target power consumption data subsequence and the second target power consumption data subsequence is greater than a preset deviation threshold, wherein the first target power consumption data subsequence and the second target power consumption data subsequence are two adjacent target power consumption data subsequences; if the sequence deviation degree is greater than the preset deviation threshold, define the first target power consumption data subsequence and the second target power consumption data subsequence as a first suspicious power consumption data subsequence and a second suspicious power consumption data subsequence respectively, and perform power consumption data anomaly analysis on the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence according to a preset first data analysis strategy; perform abnormal power consumption data anomaly analysis on a first non-target power consumption data subsequence adjacent to the first suspicious power consumption data subsequence and a second non-target power consumption data subsequence adjacent to the second suspicious power consumption data subsequence according to a preset second data analysis strategy based on the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence.

[0006] In a second aspect, the present application provides a power consumption data anomaly analysis system, comprising: an acquisition module configured to acquire a power consumption data sequence and a device temperature data sequence of a power consumption device within a preset time period, wherein the power consumption data sequence contains power consumption data at different time nodes, and the device temperature data sequence contains device temperature data at different time nodes; a cutting module configured to select at least one abrupt device temperature data in the device temperature data sequence according to a change trend of each device temperature data, and cut at least one target power consumption data subsequence and at least one non-target power consumption data subsequence from the power consumption data sequence aligned with the device temperature data sequence based on a preset data cutting rule according to a time node of the at least one abrupt device temperature data; a determination module configured to determine whether a sequence deviation degree between a first target power consumption data subsequence and a second target power consumption data subsequence is greater than a preset deviation threshold, wherein the first target power consumption data subsequence and the second target power consumption data subsequence are two adjacent target power consumption data subsequences; a first analysis module configured to, if the sequence deviation degree is greater than the preset deviation threshold, define the first target power consumption data subsequence and the second target power consumption data subsequence as a first suspicious power consumption data subsequence and a second suspicious power consumption data subsequence respectively, and perform power consumption data anomaly analysis on the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence according to a preset first data analysis strategy; The second analysis module is configured to perform abnormal power consumption data analysis on a first non-target power consumption data subsequence adjacent to the first suspicious power consumption data subsequence and a second non-target power consumption data subsequence adjacent to the second suspicious power consumption data subsequence according to the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence and by using a preset second data analysis strategy.

[0007] In a third aspect, an electronic device is provided, which includes at least one processor and a memory connected to the at least one processor in communication, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the steps of the power consumption data anomaly analysis method of any of the embodiments of the present application.

[0008] In a fourth aspect, the present application further provides a computer readable storage medium having a computer program stored thereon, and the program instructions are executed by a processor to enable the processor to perform the steps of the power consumption data anomaly analysis method of any of the embodiments of the present application.

[0009] The power consumption data anomaly analysis method and system of the present application determine temperature mutation points according to dynamically preset temperature thresholds, intercept power consumption key windows according to the temperature mutation points, and adaptively focus on abnormal power consumption correlation periods. Then, the power consumption quantity trajectory form difference is quantified by sequence deviation degree, and normal load fluctuation and real failure are distinguished as much as possible. In addition, the first data analysis strategy is combined to implement accurate positioning in the suspicious area and the non-target area, i.e., dynamically selecting global marking or selective marking mode according to the maximum power consumption change, so as to effectively speed up the power consumption data anomaly identification speed on the premise of improving the accuracy of abnormal power consumption data identification as much as possible. BRIEF DESCRIPTION OF DRAWINGS

[0010] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0011] Figure 1 A flowchart of a power consumption data anomaly analysis method provided by an embodiment of the present application; Figure 2 A structural block diagram of a power consumption data anomaly analysis system provided by an embodiment of the present application; Figure 3 A structural schematic diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0012] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0013] Please refer to Figure 1 , which shows a flowchart of a power consumption data anomaly analysis method of the present application.

[0014] As Figure 1 shown, the power consumption data anomaly analysis method specifically includes the following steps: Step S101, obtaining a power consumption data sequence and a device temperature data sequence of the power consumption equipment in a preset time period, wherein the power consumption data sequence contains power consumption data of different time nodes, and the device temperature data sequence contains device temperature data of different time nodes.

[0015] In this step, a high-precision smart power meter (such as 0.5S level precision) is installed in the power supply circuit of the power consumption equipment, supporting millisecond-level current / power sampling. The power consumption data of the power consumption equipment in the preset time period is obtained through the high-precision smart power meter, and the power consumption data sequence is obtained based on the chronological order.

[0016] In the key heating parts of the equipment (such as motor winding, wiring terminal), infrared temperature sensors (precision ±1℃) or optical fiber temperature sensors (anti-electromagnetic interference) are deployed. The sensor distribution points need to cover the joule law hot spot area (such as high incidence of poor contact). The device temperature data of the power consumption equipment in the preset time period is obtained through the temperature sensor, and the device temperature data sequence is obtained based on the chronological order.

[0017] Step S102, selecting at least one abrupt device temperature data in the device temperature data sequence according to the change trend of each device temperature data, and according to the time node of the at least one abrupt device temperature data, obtaining at least one target power consumption data subsequence and at least one non-target power consumption data subsequence in the power consumption data sequence aligned with the device temperature data sequence based on a preset data interception rule.

[0018] In this step, the dynamic preset temperature threshold is constructed, specifically including: setting an initial preset temperature threshold, the initial preset temperature threshold corresponding to a first temperature difference value, wherein the first temperature difference value is the difference between a first device temperature data at a first position and a second device temperature data at a second position in the device temperature data sequence; determining whether the first temperature difference value is greater than the initial preset temperature threshold; if the first temperature difference value is greater than the initial preset temperature threshold, then adding the first temperature difference value and the initial preset temperature threshold to obtain an updated first preset temperature threshold, wherein the first preset temperature threshold corresponds to a second temperature difference value, the second temperature difference value being the difference between the second device temperature data at the second position and a third device temperature data at a third position in the device temperature data sequence; determining whether the second temperature difference value is greater than the first preset temperature threshold; if the second temperature difference value is greater than the first preset temperature threshold, then adding the second temperature difference value and the initial preset temperature threshold to obtain an updated second preset temperature threshold, wherein the second preset temperature threshold corresponds to a third temperature difference value, the third temperature difference value being the difference between the third device temperature data at the third position and a fourth device temperature data at a fourth position in the device temperature data sequence.

[0019] In one specific embodiment, after determining whether the first temperature difference value is greater than the initial preset temperature threshold, if the first temperature difference value is not greater than the initial preset temperature threshold, then directly defining the initial preset temperature threshold as the first preset temperature threshold; if the second temperature difference value is not greater than the first preset temperature threshold, then directly defining the first preset temperature threshold as the second preset temperature threshold.

[0020] In one specific application scenario, the device is: a transformer of a certain substation; Data acquisition frequency: 1 Hz (1 data point per second); Initial preset temperature threshold: 5°C (set according to historical data); Temperature data sequence is (66, 67, 68, 70, 78, 84, 95, 103); The first temperature difference value is ΔT1 = T1 - T0 = 67 - 66 = 1°C, since the first temperature difference value 1°C is not greater than the initial preset temperature threshold, then the initial preset temperature threshold (5°C) is defined as the first preset temperature threshold; The second temperature difference value is ΔT2 = T2 - T1 = 68 - 67 = 1°C, since the second temperature difference value 1°C is not greater than the first preset temperature threshold, then the first preset temperature threshold (5°C) is defined as the second preset temperature threshold; The third temperature difference value is ΔT3 = T3 - T2 = 70 - 68 = 2°C, since the third temperature difference value 2°C is not greater than the second preset temperature threshold, then the second preset temperature threshold (5°C) is defined as the third preset temperature threshold; The fourth temperature difference is AT4 = T4 - T3 = 78 - 70 = 8 °C, and since the fourth temperature difference 8 °C is greater than the third preset temperature threshold, the third temperature difference (2 °C) is added to the initial preset temperature threshold (5 °C) to obtain a fifth preset temperature threshold (7 °C); The fifth temperature difference is AT5 = T5 - T4 = 84 - 78 = 6 °C, and since the fifth temperature difference 6 °C is not greater than the fifth preset temperature threshold (7 °C), the fifth preset temperature threshold is defined as a sixth preset temperature threshold. The sixth temperature difference is AT6 = T6 - T5 = 95 - 84 = 11 °C, and since the sixth temperature difference 11 °C is greater than the sixth preset temperature threshold (7 °C), the third temperature difference (6 °C) is added to the initial preset temperature threshold (5 °C) to obtain a seventh preset temperature threshold (11 °C). The seventh temperature difference is AT7 = T7 - T6 = 103 - 95 = 8 °C, and since the seventh temperature difference 8 °C is not greater than the seventh preset temperature threshold (11 °C), the seventh preset temperature threshold is defined as an eighth preset temperature threshold.

[0021] Further, the temperature difference between each two adjacent device temperature data in the device temperature data sequence is compared with the corresponding dynamic preset temperature threshold, and whether there is at least one abrupt device temperature data in the device temperature data sequence is determined according to the comparison result; if there is no at least one abrupt device temperature data, each power consumption data in the power consumption data sequence corresponding to the device temperature data sequence is directly determined to be normal; if there is at least one abrupt device temperature data, at least one abrupt device temperature data is selected from the device temperature data sequence.

[0022] In one specific embodiment, the temperature data sequence is (66, 67, 68, 70, 78, 84, 95, 103), wherein the difference between 78 °C and 70 °C is greater than the corresponding third preset temperature threshold, so 78 °C is defined as an abrupt device temperature data, and the difference between 95 °C and 84 °C is greater than the corresponding sixth preset temperature threshold, so 95 °C is also defined as an abrupt device temperature data. Therefore, the first failure and the second failure may occur at the time points corresponding to 78 °C and 95 °C, respectively.

[0023] In this embodiment, by setting a dynamic preset temperature threshold, the subsequent temperature rise can be made more sensitive, so as to capture as accurately as possible the starting point of the temperature rise which is not caused by the normal load work of the electrical equipment. Since there is a lag between the temperature change and the power change when a fault occurs, the time point corresponding to the sudden change in equipment temperature data cannot be directly defined as the occurrence time point of the power anomaly. Therefore, it is necessary to perform the step of "obtaining at least one target power data subsequence and at least one non-target power data subsequence from the power data sequence aligned with the equipment temperature data sequence based on a preset data extraction rule" and other subsequent steps to further analyze the time point of the power data anomaly, so that this embodiment can achieve the effect of quickly determining the power data anomaly time interval.

[0024] It should be noted that, based on the time node of at least one sudden change equipment temperature data, at least one target power data subsequence and at least one non-target power data subsequence are obtained from the power data sequence aligned with the equipment temperature data sequence based on a preset data extraction rule, including: aligning the equipment temperature data sequence with the power data sequence, selecting power data at the same time node as the at least one sudden change equipment temperature data from the power data sequence, and defining the power data as sudden change power data; taking a position point of a certain sudden change power data as a center point, and extracting a certain power data subsequence of a preset length from the power data sequence; judging whether the absolute difference of power between two adjacent power data in a certain power data subsequence is greater than a preset power threshold based on the time sequence; if the absolute difference of power between any two adjacent power data in a certain power data subsequence is not greater than the preset power threshold, the certain power data subsequence is directly defined as a target power data subsequence; if the absolute difference of power between the first power data and the second power data in a certain power data subsequence is greater than the preset power threshold, the position point of the first power data is defined as a first cut-off point, wherein the time node corresponding to the first power data is earlier than the time node corresponding to the second power data, and the first power data and the second power data are the first two adjacent power data greater than the preset power threshold based on the time sequence; taking the center point as a symmetric point to obtain a second cut-off point symmetric to the first cut-off point, and extracting the power data between the first cut-off point and the second cut-off point in a certain power data subsequence to obtain a target power data subsequence; extracting each target power data subsequence from the power data sequence to obtain at least one target power data subsequence and at least one non-target power data subsequence.

[0025] For example, the power data at the time points corresponding to 78℃ and 95℃ is obtained, and the power data at the two time points is defined as sudden change power data.

[0026] In the embodiment, the key window of the power consumption data is anchored by the temperature mutation point (e.g., 78℃, 95℃), the steady-state data interference is excluded, and the failure correlation period is focused on. In addition, the analysis window is automatically adjusted by the symmetric interception mechanism: when the power consumption mutation is significant (> preset power consumption threshold), the analysis range is reduced to the mutation interval; when the power consumption changes gently (≤ preset power consumption threshold), the complete preset window is retained to avoid the problem that the preset window length does not match the failure characteristics.

[0027] In step S103, it is determined whether the sequence deviation degree between the first target power consumption data subsequence and the second target power consumption data subsequence is greater than a preset deviation threshold, wherein the first target power consumption data subsequence and the second target power consumption data subsequence are two adjacent target power consumption data subsequences.

[0028] In this step, before determining whether the sequence deviation degree between the first target power consumption data subsequence and the second target power consumption data subsequence is greater than a preset deviation threshold, the first target power consumption data subsequence and the second target power consumption data subsequence are aligned based on the center points of the target power consumption data subsequences; the target power consumption data at the first center point of the first target power consumption data subsequence is defined as the first center target power consumption data, and the target power consumption data at the second center point of the second target power consumption data subsequence is defined as the first center target power consumption data; the first power consumption absolute difference value between the first other target power consumption data and the first center target power consumption data is obtained, wherein the first other target power consumption data is the first target power consumption data in the first target power consumption data subsequence except the first center target power consumption data; the second power consumption absolute difference value between the second other target power consumption data and the second center target power consumption data is obtained, wherein the second other target power consumption data is the second target power consumption data in the second target power consumption data subsequence except the second center target power consumption data; the first power consumption absolute difference value of the first other target power consumption data and the second power consumption absolute difference value of the second other target power consumption data at the same position after alignment are subtracted again, and the deviation degree of each power consumption absolute difference value is obtained by taking the absolute value, and the deviation degrees are added to obtain the sequence deviation degree between the first target power consumption data subsequence and the second target power consumption data subsequence.

[0029] Specifically, the power consumption absolute difference value is the absolute value of the power consumption difference value, i.e., the first power consumption absolute difference value is the absolute value of the first power consumption difference value, and the second power consumption absolute difference value is the absolute value of the second power consumption difference value.

[0030] In one specific application scenario, the power consumption data is the power consumption, and the unit is kW·h. The first target power consumption data subsequence is [475, 480, 483, 485 (center), 492, 505, 510], and the first center point is 485. The second target power consumption data subsequence is [505, 520, 545, 565 (center), 575, 580, 578], and the second center point is 565. The first center point 485 of the first target power consumption data subsequence is aligned with the second center point 565 of the second target power consumption data subsequence on the time axis. First difference (difference between each point and the center point): The difference result of the first target power consumption data subsequence is [10, 5, 2, 7, 20, 25]. The difference result of the first target power consumption data subsequence is [60, 45, 20, 10, 15, 13]. Second difference (second difference at the aligned position): Position 1: 60-10 = 50. Position 2: 45-5 = 40. Position 3: 20-2 = 18. Position 4: 10-7 = 3. Position 5: 20-15 = 5. Position 6: 25-13 = 12.

[0031] Sequence deviation degree = 50+40+18+3+5+12 = 128.

[0032] In this embodiment, by judging whether the sequence deviation degree of adjacent target power consumption data subsequences exceeds the preset deviation threshold, normal load fluctuations (such as smooth and symmetric trajectories of motor start and stop) and real faults (such as power consumption mutations caused by faults) can be accurately distinguished, which facilitates further abnormal analysis of power consumption data in subsequent steps S104 and S105.

[0033] In step S104, if the deviation degree is greater than the preset deviation threshold, the first target power consumption data subsequence and the second target power consumption data subsequence are defined as a first suspicious power consumption data subsequence and a second suspicious power consumption data subsequence, respectively, and power consumption data abnormal analysis is performed on the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence according to a preset first data analysis strategy.

[0034] In this step, the first maximum power consumption change quantity in the first suspicious power consumption data subsequence and the second maximum power consumption change quantity in the second suspicious power consumption data subsequence are obtained, and it is judged whether the first maximum power consumption change quantity and the second maximum power consumption change quantity are greater than a preset power consumption change quantity threshold; if the first maximum power consumption change quantity is not greater than the preset power consumption change quantity threshold, each first suspicious power consumption data in the first suspicious power consumption data subsequence is defined as abnormal power consumption data, otherwise, the first suspicious power consumption data greater than the first suspicious average power consumption quantity value in the first suspicious power consumption data subsequence is defined as abnormal power consumption data; if the second maximum power consumption change quantity is not greater than the preset power consumption change quantity threshold, each second suspicious power consumption data in the second suspicious power consumption data subsequence is defined as abnormal power consumption data, otherwise, the second suspicious power consumption data greater than the second suspicious average power consumption quantity value in the second suspicious power consumption data subsequence is defined as abnormal power consumption data.

[0035] In this embodiment, the first maximum power consumption change quantity is greater than the preset power consumption change quantity threshold, so that a larger change in power consumption occurs in the first suspicious power consumption data subsequence, and thus it is possible that the time when the fault point occurs is in the time interval corresponding to the first suspicious power consumption data subsequence with a high probability, and thus the first suspicious power consumption data greater than the first suspicious average power consumption quantity value in the first suspicious power consumption data subsequence is defined as abnormal power consumption data, wherein the first suspicious average power consumption quantity value is the average of each first suspicious power consumption data in the first suspicious power consumption data subsequence; when the first maximum power consumption change quantity is not greater than the preset power consumption change quantity threshold, it is possible that the time when the fault point occurs is not in the time interval corresponding to the first suspicious power consumption data subsequence with a high probability, but the fault has occurred before the time interval corresponding to the first suspicious power consumption data subsequence, resulting in that each first suspicious power consumption data in the first suspicious power consumption data subsequence is continuously abnormal.

[0036] When the maximum power consumption change quantity of the suspicious subsequence does not exceed the preset power consumption change quantity threshold (indicating that the anomaly is a persistent and gradual accumulation), the entire suspicious subsequence is determined as a global anomaly to avoid local missed detection of gradual faults; and when the maximum change quantity exceeds the preset power consumption change quantity threshold (representing a sudden and severe fluctuation), only the data points higher than the average power consumption quantity are defined as anomalies, effectively eliminating normal fluctuation sections. This strategy can accurately realize pixel-level positioning of the abnormal power consumption data anomaly time by dynamically switching between "overall marking" and "selective marking" modes.

[0037] In step S105, according to the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence, a preset second data analysis strategy is used to perform abnormal power consumption data anomaly analysis on a first non-target power consumption data subsequence adjacent to the first suspicious power consumption data subsequence and a second non-target power consumption data subsequence adjacent to the second suspicious power consumption data subsequence.

[0038] In this step, when the first maximum power consumption change amount is greater than the preset power consumption change amount threshold, it is directly determined that each first non-target power consumption data in the first non-target power consumption data subsequence is normal power consumption data. When the first maximum power consumption change amount is not greater than the preset power consumption change amount threshold, the first non-target power consumption data in the first non-target power consumption data subsequence greater than the first non-target average power consumption amount is defined as abnormal power consumption data. When the second maximum power consumption change amount is greater than the preset power consumption change amount threshold, it is directly determined that each second non-target power consumption data in the second non-target power consumption data subsequence is normal power consumption data. When the second maximum power consumption change amount is not greater than the preset power consumption change amount threshold, the second non-target power consumption data in the second non-target power consumption data subsequence greater than the second non-target average power consumption amount is defined as abnormal power consumption data.

[0039] In this embodiment, the second data analysis strategy is adopted, which can as much as possible realize accurate diagnosis of the non-target power consumption data subsequence: when the maximum power consumption change amount of the suspicious subsequence exceeds the preset threshold (indicating a sudden local fault), it is determined that the adjacent non-target subsequence is normal as a whole, which can better diagnose the non-anomalous time interval; and when the maximum change amount of the suspicious subsequence does not exceed the threshold (indicating a persistent hidden fault), the data points in the non-target subsequence higher than the average power consumption are marked as abnormal, so that the abnormal power consumption data can be accurately determined.

[0040] In summary, the method of the present application determines the temperature mutation point according to the dynamically preset temperature threshold, intercepts the power consumption key window according to the temperature mutation point, and adaptively focuses on the abnormal power consumption correlation period; then the power consumption trajectory form difference is quantified by the sequence deviation degree, so as to distinguish the normal load fluctuation and the real fault as much as possible; and the suspicious area and the non-target area are accurately positioned by combining the first data analysis strategy, that is, the global marking or selective marking mode is dynamically selected according to the maximum power consumption change amount, so that the speed of power consumption data anomaly identification is effectively accelerated on the premise of as much as possible improving the accuracy of abnormal power consumption data identification.

[0041] Please refer to Figure 2 which shows the structure block diagram of a power consumption data anomaly analysis system of the present application.

[0042] As Figure 2 shown, the power consumption data anomaly analysis system 200 includes an acquisition module 210, an interception module 220, a judgment module 230, a first analysis module 240, and a second analysis module 250.

[0043] The acquisition module 210 is configured to acquire a power consumption data sequence and a device temperature data sequence of the power consumption device in a preset time period, wherein the power consumption data sequence comprises power consumption data at different time nodes, and the device temperature data sequence comprises device temperature data at different time nodes. The intercepting module 220 is configured to select at least one abrupt device temperature data in the device temperature data sequence according to a change trend of each device temperature data, and according to a time node of the at least one abrupt device temperature data, obtain at least one target power consumption data subsequence and at least one non-target power consumption data subsequence in the power consumption data sequence aligned with the device temperature data sequence based on a preset data intercepting rule. The judging module 230 is configured to judge whether a sequence deviation degree between a first target power consumption data subsequence and a second target power consumption data subsequence is greater than a preset deviation threshold, wherein the first target power consumption data subsequence and the second target power consumption data subsequence are two adjacent target power consumption data subsequences. The first analyzing module 240 is configured to, if the sequence deviation degree is greater than the preset deviation threshold, define the first target power consumption data subsequence and the second target power consumption data subsequence as a first suspicious power consumption data subsequence and a second suspicious power consumption data subsequence respectively, and perform power consumption data anomaly analysis on the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence according to a preset first data analysis strategy. The second analyzing module 250 is configured to perform abnormal power consumption data anomaly analysis on a first non-target power consumption data subsequence adjacent to the first suspicious power consumption data subsequence and a second non-target power consumption data subsequence adjacent to the second suspicious power consumption data subsequence according to the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence, and using a preset second data analysis strategy.

[0044] It should be understood that Figure 2 the modules described in the above Figure 1 correspond to the respective steps in the methods described in the above Figure 2 The operations and features described above for the methods also apply to the modules in the above

[0045] In some other embodiments, the present application also provides a computer readable storage medium having stored thereon a computer program, the program instructing a processor to execute the power consumption data anomaly analysis method in any of the above method embodiments when the program is executed by the processor. As an implementation form, the computer readable storage medium of the present application stores computer executable instructions, and the computer executable instructions are configured to: obtain a power consumption data sequence and a device temperature data sequence of the power consumption device in a preset time period, wherein the power consumption data sequence comprises power consumption data at different time nodes, and the device temperature data sequence comprises device temperature data at different time nodes; select at least one abrupt device temperature data in the device temperature data sequence according to a change trend of each device temperature data, and obtain at least one target power consumption data subsequence and at least one non-target power consumption data subsequence in the power consumption data sequence after alignment with the device temperature data sequence based on a preset data interception rule according to a time node of the at least one abrupt device temperature data; determine whether a sequence deviation degree between a first target power consumption data subsequence and a second target power consumption data subsequence is greater than a preset deviation threshold, wherein the first target power consumption data subsequence and the second target power consumption data subsequence are two adjacent target power consumption data subsequences; if greater than the preset deviation threshold, define the first target power consumption data subsequence and the second target power consumption data subsequence as a first suspicious power consumption data subsequence and a second suspicious power consumption data subsequence respectively, and perform power consumption data anomaly analysis on the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence according to a preset first data analysis strategy; perform abnormal power consumption data anomaly analysis on a first non-target power consumption data subsequence adjacent to the first suspicious power consumption data subsequence and a second non-target power consumption data subsequence adjacent to the second suspicious power consumption data subsequence according to a preset second data analysis strategy.

[0046] The computer readable storage medium can include a program storage area and a data storage area, wherein the program storage area can store an operating system and at least one application required by a function; and the data storage area can store data created according to use of the power consumption data anomaly analysis system. In addition, the computer readable storage medium can include a high-speed random access memory, and can also include a memory such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state memory device. In some embodiments, the computer readable storage medium can optionally include a memory remotely arranged relative to the processor, and these remote memories can be connected to the power consumption data anomaly analysis system through a network. Examples of the above network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0047] Figure 3 is a structural schematic diagram of an electronic device provided by an embodiment of the application, as Figure 3As shown, the device includes a processor 310 and a memory 320. The electronic device can also include an input device 330 and an output device 340. The processor 310, the memory 320, the input device 330 and the output device 340 can be connected by a bus or other means, Figure 3 The memory 320 is the computer readable storage medium described above. The processor 310 performs various functional applications and data processing of the server by running the non-volatile software programs, instructions and modules stored in the memory 320, that is, implements the power consumption data anomaly analysis method of the method embodiment described above. The input device 330 can receive input digital or character information, and generate key signal input related to user settings and function control of the power consumption data anomaly analysis system. The output device 340 can include a display device such as a display screen.

[0048] The electronic device described above can execute the method provided by the embodiments of the present application, and has the corresponding function modules and beneficial effects of executing the method. Technical details not described in detail in the present embodiment can be referred to the method provided by the embodiments of the present application.

[0049] As an implementation, the electronic device described above is applied to the power consumption data anomaly analysis system, and is used for a client, including: at least one processor; and a memory in communication connection with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to: obtain a power consumption data sequence and a device temperature data sequence of the power consumption device within a preset time period, wherein the power consumption data sequence contains power consumption data of different time nodes, and the device temperature data sequence contains device temperature data of different time nodes; select at least one abrupt device temperature data in the device temperature data sequence according to the change trend of each device temperature data, and obtain at least one target power consumption data subsequence and at least one non-target power consumption data subsequence in the power consumption data sequence aligned with the device temperature data sequence based on a preset data interception rule according to the time node of the at least one abrupt device temperature data; determine whether the sequence deviation degree between the first target power consumption data subsequence and the second target power consumption data subsequence is greater than a preset deviation threshold, wherein the first target power consumption data subsequence and the second target power consumption data subsequence are two adjacent target power consumption data subsequences; If greater than the preset deviation threshold, the first target power consumption data subsequence and the second target power consumption data subsequence are defined as a first suspicious power consumption data subsequence and a second suspicious power consumption data subsequence respectively, and power consumption data anomaly analysis is performed on the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence according to a preset first data analysis strategy; According to the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence, a preset second data analysis strategy is adopted to perform abnormal power consumption data anomaly analysis on a first non-target power consumption data subsequence adjacent to the first suspicious power consumption data subsequence and a second non-target power consumption data subsequence adjacent to the second suspicious power consumption data subsequence.

[0050] Through the above description of the embodiments, those skilled in the art can clearly understand that the embodiments can be realized by means of software and the necessary general hardware platform, and of course, can also be realized by hardware. Based on such understanding, the above technical solutions can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods of the various embodiments or some parts of the embodiments.

[0051] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for analyzing abnormality of power consumption data, characterized by comprising: The method comprises: obtaining a power consumption data sequence and a device temperature data sequence of a power consumption device within a preset time period; selecting at least one abrupt device temperature data from the device temperature data sequence according to the change trend of each device temperature data, and obtaining at least one target power consumption data subsequence and at least one non-target power consumption data subsequence from the power consumption data sequence aligned with the device temperature data sequence based on a preset data interception rule according to the time node of the at least one abrupt device temperature data; judging whether the sequence deviation degree between the first target power consumption data subsequence and the second target power consumption data subsequence is greater than a preset deviation threshold; if greater than the preset deviation threshold, defining the first target power consumption data subsequence and the second target power consumption data subsequence as a first suspicious power consumption data subsequence and a second suspicious power consumption data subsequence respectively, and performing power consumption data anomaly analysis on the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence according to a preset first data analysis strategy; performing abnormal power consumption data anomaly analysis on the first non-target power consumption data subsequence adjacent to the first suspicious power consumption data subsequence and the second non-target power consumption data subsequence adjacent to the second suspicious power consumption data subsequence according to a preset second data analysis strategy. 2.The method of claim 1, wherein, Wherein, the power consumption data sequence contains power consumption data at different time nodes, and the device temperature data sequence contains device temperature data at different time nodes; the at least one abrupt device temperature data is selected from the device temperature data sequence according to the change trend of each device temperature data, which comprises: constructing a dynamic preset temperature threshold, specifically: setting an initial preset temperature threshold, which corresponds to a first temperature difference value, wherein the first temperature difference value is the difference between a first device temperature data at a first position and a second device temperature data at a second position in the device temperature data sequence; judging whether the first temperature difference value is greater than the initial preset temperature threshold; if the first temperature difference value is greater than the initial preset temperature threshold, adding the first temperature difference value and the initial preset temperature threshold to obtain an updated first preset temperature threshold, wherein the first preset temperature threshold corresponds to a second temperature difference value, and the second temperature difference value is the difference between the second device temperature data at the second position and a third device temperature data at a third position in the device temperature data sequence; judging whether the second temperature difference value is greater than the first preset temperature threshold; if the second temperature difference value is greater than the first preset temperature threshold, adding the second temperature difference value and the initial preset temperature threshold to obtain an updated second preset temperature threshold, wherein the second preset temperature threshold corresponds to a third temperature difference value, and the third temperature difference value is the difference between the third device temperature data at the third position and a fourth device temperature data at a fourth position in the device temperature data sequence; The temperature difference between two adjacent device temperature data in the device temperature data sequence is compared with a corresponding dynamic preset temperature threshold, and whether at least one mutation device temperature data exists in the device temperature data sequence is determined according to a comparison result; If at least one mutation device temperature data does not exist, each power consumption data in a power consumption data sequence corresponding to the device temperature data sequence is directly determined to be normal; If at least one mutation device temperature data exists, the at least one mutation device temperature data is selected in the device temperature data sequence. 3.The method of claim 2, wherein, Wherein, The first target power consumption data sub-sequence and the second target power consumption data sub-sequence are two adjacent target power consumption data sub-sequences; After determining whether the first temperature difference is greater than the initial preset temperature threshold, the method further comprises: If the first temperature difference is not greater than the initial preset temperature threshold, the initial preset temperature threshold is directly defined as a first preset temperature threshold; If the second temperature difference is not greater than the first preset temperature threshold, the first preset temperature threshold is directly defined as a second preset temperature threshold.

4. The method of claim 1, wherein, The at least one target power consumption data sub-sequence and the at least one non-target power consumption data sub-sequence obtained based on the preset data interception rule in the power consumption data sequence aligned with the device temperature data sequence according to the time node of the at least one mutation device temperature data comprises: Aligning the device temperature data sequence with the power consumption data sequence, and selecting power consumption data at the same time node as the at least one mutation device temperature data in the power consumption data sequence and defining it as a mutation power consumption data; Taking a position point where a certain mutation power consumption data is located as a center point, a certain power consumption data sub-sequence of a preset length is intercepted in the power consumption data sequence; Judging whether an absolute power consumption difference between two adjacent power consumption data in the certain power consumption data sub-sequence is greater than a preset power consumption threshold based on a time sequence; If the absolute power consumption difference between any two adjacent power consumption data in the certain power consumption data sub-sequence is not greater than the preset power consumption threshold, the certain power consumption data sub-sequence is directly defined as a target power consumption data sub-sequence; If the absolute power consumption difference between a first power consumption data and a second power consumption data in the certain power consumption data sub-sequence is greater than the preset power consumption threshold, a position point where the first power consumption data is located is defined as a first cutoff point, wherein the time node corresponding to the first power consumption data is earlier than the time node corresponding to the second power consumption data, and the first power consumption data and the second power consumption data are the first two adjacent power consumption data greater than the preset power consumption threshold based on the time sequence; Taking the center point as a symmetry point, a second cutoff point symmetric to the first cutoff point is obtained, and power consumption data between the first cutoff point and the second cutoff point in the certain power consumption data sub-sequence is intercepted to obtain a target power consumption data sub-sequence; Each target power consumption data sub-sequence is intercepted in the power consumption data sequence to obtain at least one target power consumption data sub-sequence and at least one non-target power consumption data sub-sequence.

5. The method of claim 1, wherein, Before judging whether the sequence deviation degree between the first target power consumption data subsequence and the second target power consumption data subsequence is greater than a preset deviation threshold, the method further comprises: aligning the first target power consumption data subsequence and the second target power consumption data subsequence based on the center points of the target power consumption data subsequences; defining the target power consumption data at the first center point of the first target power consumption data subsequence as first center target power consumption data, and the target power consumption data at the second center point of the second target power consumption data subsequence as first center target power consumption data; obtaining a first power consumption absolute difference value between the first other target power consumption data and the first center target power consumption data, wherein the first other target power consumption data is the first target power consumption data in the first target power consumption data subsequence except the first center target power consumption data; obtaining a second power consumption absolute difference value between the second other target power consumption data and the second center target power consumption data, wherein the second other target power consumption data is the second target power consumption data in the second target power consumption data subsequence except the second center target power consumption data; subtracting the first power consumption absolute difference value of the first other target power consumption data from the second power consumption absolute difference value of the second other target power consumption data to obtain a deviation degree of each power consumption absolute difference value, and adding each deviation degree to obtain the sequence deviation degree between the first target power consumption data subsequence and the second target power consumption data subsequence.

6. The method of claim 1, wherein, After judging whether the sequence deviation degree between the first target power consumption data subsequence and the second target power consumption data subsequence is greater than a preset deviation threshold, the method further comprises: if not greater than the preset deviation threshold, performing power consumption data anomaly analysis on the first target power consumption data subsequence according to the first data analysis strategy, wherein the sequence length of the first target power consumption data subsequence is greater than the sequence length of the second target power consumption data subsequence; assigning the analysis result of each first target power consumption data in the first target power consumption data subsequence to each second target power consumption data in the same position in the second target power consumption data subsequence.

7. The method of claim 1, wherein the method further comprises: The power consumption data anomaly analysis on the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence according to the preset first data analysis strategy comprises: obtaining a first maximum power consumption change value in the first suspicious power consumption data subsequence and a second maximum power consumption change value in the second suspicious power consumption data subsequence, and judging whether the first maximum power consumption change value and the second maximum power consumption change value are greater than a preset power consumption change threshold; if the first maximum power consumption change value is not greater than the preset power consumption change threshold, defining each first suspicious power consumption data in the first suspicious power consumption data subsequence as abnormal power consumption data, otherwise defining the first suspicious power consumption data greater than the first suspicious average power consumption value in the first suspicious power consumption data subsequence as abnormal power consumption data; If the second maximum power consumption change amount is not greater than the preset power consumption change amount threshold, each second suspicious power consumption data in the second suspicious power consumption data subsequence is defined as abnormal power consumption data, otherwise, the second suspicious power consumption data greater than the second suspicious average power consumption amount in the second suspicious power consumption data subsequence is defined as abnormal power consumption data.

8. The method of claim 7, wherein the method further comprises: The second data analysis strategy is used to analyze the first non-target power consumption data subsequence adjacent to the first suspicious power consumption data subsequence and the second non-target power consumption data subsequence adjacent to the second suspicious power consumption data subsequence according to the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence. When the first maximum power consumption change amount is greater than the preset power consumption change amount threshold, each first non-target power consumption data in the first non-target power consumption data subsequence is directly determined as normal power consumption data. When the first maximum power consumption change amount is not greater than the preset power consumption change amount threshold, the first non-target power consumption data greater than the first non-target average power consumption amount in the first non-target power consumption data subsequence is defined as abnormal power consumption data. When the second maximum power consumption change amount is greater than the preset power consumption change amount threshold, each second non-target power consumption data in the second non-target power consumption data subsequence is directly determined as normal power consumption data. When the second maximum power consumption change amount is not greater than the preset power consumption change amount threshold, the second non-target power consumption data greater than the second non-target average power consumption amount in the second non-target power consumption data subsequence is defined as abnormal power consumption data.

9. A power usage data anomaly analysis system, comprising: The method comprises: An acquisition module configured to acquire power consumption data sequences and device temperature data sequences of power consumption devices within a preset time period; An intercepting module configured to select at least one abrupt device temperature data in the device temperature data sequence according to the change trend of each device temperature data, and based on the time node of the at least one abrupt device temperature data, obtain at least one target power consumption data subsequence and at least one non-target power consumption data subsequence in the power consumption data sequence aligned with the device temperature data sequence according to a preset data intercepting rule; A judging module configured to judge whether the sequence deviation degree between the first target power consumption data subsequence and the second target power consumption data subsequence is greater than a preset deviation threshold; A first analysis module configured to, if greater than the preset deviation threshold, define the first target power consumption data subsequence and the second target power consumption data subsequence as a first suspicious power consumption data subsequence and a second suspicious power consumption data subsequence respectively, and perform power consumption data anomaly analysis on the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence according to a preset first data analysis strategy. The second analysis module is configured to perform abnormal power consumption data analysis on a first non-target power consumption data subsequence adjacent to the first suspicious power consumption data subsequence and a second non-target power consumption data subsequence adjacent to the second suspicious power consumption data subsequence according to the first suspicious power consumption data subsequence and the second suspicious power consumption data subsequence and by using a preset second data analysis strategy.

Images