CN-115423128-B - Method for monitoring non-invasive abnormal load behavior, electronic device and storage medium

CN115423128BCN 115423128 BCN115423128 BCN 115423128BCN-115423128-B

Abstract

The invention relates to a monitoring method of non-invasive abnormal load behaviors, electronic equipment and a storage medium, wherein the method comprises the steps of acquiring real-time monitoring data of the non-invasive load monitoring equipment and denoising, wherein the monitoring data comprise total voltage and total current data of a preselected power circuit monitored by the non-invasive load monitoring equipment, taking monitoring data with load state conversion as effective monitoring data aiming at the denoised monitoring data, carrying out color coding processing on the effective monitoring data based on a pre-built power strategy to obtain V-I track images of total voltage and total current corresponding to each power circuit, inputting the V-I track images into a training condition to generate an countermeasure network, and judging whether abnormal loads exist in each power circuit based on the generated characteristic reconstruction images. The non-invasive load monitoring system has the beneficial effects that the technical problems of low non-invasive load monitoring expansibility, low flexibility and high monitoring error in the prior art can be solved.

Inventors

HAN YINGHUA
LI KEKE
FENG HANTONG
ZHAO QIANG

Assignees

东北大学秦皇岛分校

Dates

Publication Date: 20260505
Application Date: 20220831

Claims (7)

1. A method for monitoring non-intrusive abnormal load behavior, comprising the steps of: S1, acquiring real-time monitoring data of non-invasive load monitoring equipment and denoising, wherein the monitoring data comprise total voltage and total current data of a preselected power circuit monitored by the non-invasive load monitoring equipment; s2, aiming at the denoised monitoring data, taking the monitoring data with load state conversion as effective monitoring data; s3, performing color coding processing on the effective monitoring data based on a pre-constructed power strategy to obtain a V-I track image of each single load in the power circuit; s4, inputting the V-I track image into a training condition generation countermeasure network, and judging whether each power circuit has abnormal load or not based on the generated characteristic reconstruction image; The condition generating countermeasure network comprises a condition self-encoder, a capsule network and a classifier, wherein the condition self-encoder is used for realizing the conversion of Gaussian prior probability of load in a V-I track image into Gaussian posterior probability, and the capsule network is used for realizing the compactness of the same type of characteristics near a Gaussian distribution center so that the classifier detects the load; Prior to S1, the method further comprises training the condition generating countermeasure network S0: the S0 includes: S01, acquiring a training monitoring data sample and a checking monitoring data sample for training the condition generation countermeasure network, wherein the training monitoring data sample and the checking monitoring data sample are historical monitoring total voltage and total current data of the same power circuit; s02, encoding the training monitoring data sample and the verification detection data sample based on a pre-constructed power strategy, and obtaining a training V-I track image and a verification V-I track image of each single load in the power circuit; s03, inputting the training V-I track image into the condition generation countermeasure network reconstruction to generate a characteristic reconstruction image of the training V-I track for each single load; S04, respectively inputting the verification V-I track image and the characteristic reconstruction image corresponding to each single load into a pre-constructed discriminator, and judging whether the characteristic reconstruction image is matched with the verification V-I track image; S05, adjusting training parameters of the condition generation countermeasure network, and alternately generating a characteristic reconstruction image and inputting a discrimination network so that the characteristic reconstruction image finally generated by the condition countermeasure monitoring network is matched with the check V-I track image, thereby obtaining a trained condition countermeasure monitoring network; adjusting the training parameters of the condition countermeasure monitoring network, in particular, Calculating a minimum value of training loss of the condition countermeasure monitoring network based on a pre-constructed loss function; Adjusting the condition to generate parameters of an countermeasure network by weighting calculation of the minimum value; the loss functions include feature matching loss functions, reconstruction loss functions, additional encoder loss functions, center constraint loss functions, and/or contrast loss functions; the feature matching loss function expression is: ; the f (x) is given as the input V-I trajectory x, the output of the discriminator intermediate layer; The reconstruction loss function expression is: ; The said Mu is the average intensity of the training V-I track image, delta is the standard deviation of the training V-I track image, Reconstructing covariance of the image for training the V-I track and the feature, wherein c1 and c2 are constants; the additional encoder loss function expression is: ; the z is the sampling vector characteristic output by the capsule network when training the V-I track image, Reconstructing coding features of the image for the features; The center constraint loss function expression is: L KL =d(C,sg[P y ), sg represents a stop gradient operator; The said The C is a probability capsule, and the P is Gaussian distribution of a target load cluster; The contrast loss function expression is: ; The [. Cndot. ] + is a function of the positive number of the return parameter; the formula for calculating the minimum weight is as follows: L=αl KL +βL rec +γL contr +σL enc +λL adv , the α, β, γ, σ and λ being constants.
2. The monitoring method of claim 1, wherein S01 comprises: acquiring historical monitoring data of at least one load state transition event in a power circuit monitored by non-invasive load monitoring equipment, wherein the load state transition event is a circuit load transition process caused when a single load in the preselected power circuit is turned on and/or turned off; based on a predefined event detection window, calculating a time period when a load state transition event occurs, wherein the time period specifically comprises: Calculating the total real power S t of the preselected power circuit, determining Time t of S t >S on1 ; Based on a pre-constructed event detection window, the total actual power variation is calculated and determined when t=t+tr S t+TR <S on1 , wherein R is the step length of the event detection window, S t =S t +1-S t ; If S t+TR -S t <S on2 is carried out, judging that a load state transition event occurs in the time period of t-t+TR; Collecting total voltage and total current data of T time period before and after the load state transition event occurs, and obtaining a training monitoring data sample and a checking monitoring data sample; The S on1 is a predefined load state transition event start threshold, and S on2 is a predefined load state transition event end threshold.
3. The monitoring method of claim 1, wherein S3 comprises: S30, sampling the effective monitoring data based on a pre-constructed spectrum analysis method, and obtaining the voltage and current values of each single load in the power circuit; S31, determining an active component current i a (t) and a reactive component current i f (t) of a current i (t) of each single load based on a pre-constructed Fryze power strategy; Calculating and acquiring a power factor matrix based on the active component current i a (t) and the reactive component current i f (t) The power factor is the ratio of the power of the active component current to the power of the reactive component current; The power factor matrix The expression of (2) is: ; For the total number of sampling points, P active is active power, P apparent is real power, and V rms 、I rms is the effective values of load voltage and load current respectively; s32, constructing a hue matrix of the V-I track based on a pre-constructed HSV color space for each single load And a voltage cycle matrix V; S33, connecting the power factor matrix in a standard three-dimensional coordinate system for each single load Hue matrix And the voltage period matrix V is used for acquiring the V-I track image of the single load.
4. The method of monitoring as claimed in claim 3, wherein, The step S32 specifically comprises the following steps; S321, acquiring a motion direction H j of the V-I track by using a hue attribute hue based on the HSV color space; Based on the motion direction H j , the hue of the jth sampling point is stored in a matrix of 2N multiplied by 2N to obtain a hue matrix ; The motion direction H j has a calculation expression as follows: ; The said Arctangent function for four quadrants; the hue matrix The computational expression is: ; The A| is the cardinality of the collection; S322, averaging M periods of a single load voltage based on a pre-constructed binary image W m (1, 2.. M) to obtain a voltage period matrix V; The expression of the voltage period matrix V is as follows: 。
5. the method of monitoring as claimed in claim 1, wherein, The step S4 specifically comprises the following steps: inputting a real-time V-I track of the power circuit to the trained condition generation countermeasure network to generate a real-time characteristic reconstruction image; Calculating the minimum distance between the real-time characteristic reconstruction image and the final training condition to generate a historical characteristic reconstruction image of the countermeasure network; And if the minimum distance between the real-time characteristic reconstruction image and the historical characteristic reconstruction image is larger than a threshold tau meeting the preset requirement, judging that abnormal load occurs in the preselected power circuit.
6. An electronic device comprising a memory and a processor, the memory having stored therein a computer program, the processor executing the computer program stored in the memory, the processor implementing the steps of the method for monitoring non-invasive abnormal load behavior according to any of the preceding claims 1 to 5.
7. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a computer program which, when executed by a processor, implements the steps of the method for monitoring non-invasive abnormal load behaviour according to any of the preceding claims 1 to 5.

Description

Method for monitoring non-invasive abnormal load behavior, electronic device and storage medium Technical Field The present invention relates to the field of load monitoring technologies, and in particular, to a method for monitoring non-invasive abnormal load behaviors, an electronic device, and a storage medium. Background In modern society production and life, a large amount of renewable energy is generated and utilized at the consumer side, and user behavior can promote efficient integration of distributed energy sources highly dependent on weather. Thus, it is critical to observe the activity and actions of the user's electricity usage. It is particularly important to obtain real-time power consumption information of each electric appliance in the user. Different from the current intelligent ammeter and the obtained electricity consumption information of the total load, the load electricity consumption detail monitoring is to obtain the real-time electricity consumption information of each electric appliance in the power consumer through a certain technical means, wherein the real-time electricity consumption information comprises the working state, the electricity consumption power, the accumulated electric quantity, fault information and the like of the electric appliances. In the prior art, load monitoring mainly comprises an invasive type and a non-invasive type, wherein the invasive type load monitoring needs to invade the inside of a power load, a data measurement sensor with a communication function is respectively installed for each electric appliance, and then power utilization information is collected and sent out locally. To achieve the same purpose, the non-invasive load monitoring can obtain the electricity utilization information of each electric appliance in the user by analyzing the total load data only by installing a data measurement sensor with a communication function at the power supply inlet of the electric load. Compared to invasive methods, non-invasive methods are low cost, easy to install, and well-applied by detailed data obtained from non-invasive load monitoring. The existing non-invasive load monitoring is only suitable for an ideal state that the type of the electric load used by the power consumer is fixed and does not change due to the fact that an algorithm is solidified, and when electric equipment is changed or aged, larger monitoring errors can be generated, and the defects of inflexibility and inextensibility exist. Disclosure of Invention First, the technical problem to be solved In view of the above-mentioned shortcomings and disadvantages of the prior art, the present invention provides a method for monitoring non-invasive abnormal load behaviors, an electronic device and a storage medium, where the method solves the technical problems of low expansibility, low flexibility and high monitoring error of non-invasive load monitoring in the prior art. (II) technical scheme In order to achieve the above object, in a first aspect, the present invention provides a method for monitoring non-invasive abnormal load behavior, comprising the steps of: S1, acquiring real-time monitoring data of non-invasive load monitoring equipment and denoising, wherein the monitoring data comprise total voltage and total current data of a preselected power circuit monitored by the non-invasive load monitoring equipment; s2, aiming at the denoised monitoring data, taking the monitoring data with load state conversion as effective monitoring data; s3, performing color coding processing on the effective monitoring data based on a pre-constructed power strategy to obtain a V-I track image of each single load in the power circuit; s4, inputting the V-I track image into a training condition generation countermeasure network, and judging whether each power circuit has abnormal load or not based on the generated characteristic reconstruction image; The condition generating countermeasure network comprises a condition self-encoder, a capsule network and a classifier, wherein the condition self-encoder is used for realizing the conversion of Gaussian prior probability of load in a V-I track image into Gaussian posterior probability, and the capsule network is used for realizing the compactness of the same type of characteristics near the Gaussian distribution center so that the classifier detects the load. Optionally, before S1, the method further comprises S0 training the condition generating countermeasure network: the S0 includes: S01, acquiring a training monitoring data sample and a checking monitoring data sample for training the condition generation countermeasure network, wherein the training monitoring data sample and the checking monitoring data sample are historical monitoring total voltage and total current data of the same power circuit; s02, encoding the training monitoring data sample and the verification detection data sample based on a pre-constructed power strategy, and obtaining