Electric vehicle private connection charging identification method, device and equipment, storage medium and program product

Abstract

The invention relates to an electric vehicle private connection charging identification method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: acquiring total power consumption information of a user side in a power system in a preset period; the total electricity utilization information is matched with a plurality of preset charging feature libraries for a plurality of different charging stages of the electric vehicle, and stage feature matching results corresponding to the charging stages are obtained; fusing the multiple stage feature matching results to obtain a target matching result; and according to a target matching result, carrying out electric vehicle private connection charging identification. By adopting the method, more accurate non-intrusive identification of private charging of the electric vehicle can be realized.

Description

Technical Field

[0001] The present application relates to the field of power technology, and particularly to a method, device, computer device, computer-readable storage medium, and computer program product for identifying unauthorized electric vehicle charging. Background Art

[0002] The charging behavior of electric vehicles is characterized by high power, randomness, and long-term fluctuations. Its disorderly access may lead to problems such as local grid overload and three-phase imbalance, posing a severe challenge to the safe and stable operation of the distribution network. Since the behavior of users connecting electric vehicle charging equipment privately often lacks effective supervision, this exacerbates the peak-valley difference and line loss of the power grid. Therefore, how to identify whether there is unauthorized electric vehicle charging in the power system is a technical problem to be solved. Summary of the Invention

[0003] Based on this, in view of the above technical problems, it is necessary to provide a method, device, computer device, computer-readable storage medium, and computer program product for identifying unauthorized electric vehicle charging to identify whether there is unauthorized electric vehicle charging in the power system.

[0004] In a first aspect, the present application provides a method for identifying unauthorized electric vehicle charging, including:[[]]END]]

[0005] Obtaining the total electricity consumption information of the user side in the power system within a preset period;

[0006] Matching the total electricity consumption information with multiple charging feature libraries preset for multiple different charging stages of electric vehicles respectively to obtain stage feature matching results corresponding to each charging stage;

[0007] Fusing the multiple stage feature matching results to obtain a target matching result;

[0008] Identifying unauthorized electric vehicle charging according to the target matching result.

[0009] In one embodiment, the total electricity consumption information includes power information; the charging feature libraries include the time-domain features and frequency-domain features of the charging power corresponding to each charging stage;

[0010] Matching the total electricity consumption information with the charging feature libraries preset for different charging stages of electric vehicles to obtain stage feature matching results corresponding to each charging stage, including:

[0011] For the charging feature libraries corresponding to each charging stage, matching the power information with the time-domain features to obtain a time-domain matching result; matching the power information with the frequency-domain features to obtain a frequency-domain matching result;

[0012] Determine the stage feature matching result corresponding to the charging stage according to the time-domain matching result and the frequency-domain matching result.

[0013] In one embodiment, match the power information with the time-domain features to obtain the time-domain matching result, including:

[0014] Determine the pulse period, duty cycle, and pulse amplitude in the power information;

[0015] Match the power information with the time-domain features according to the pulse period, duty cycle, and pulse amplitude to obtain the time-domain matching result.

[0016] In one embodiment, match the power information with the frequency-domain features to obtain the frequency-domain matching result, including:

[0017] Match the power information with the frequency-domain features according to the short-time Fourier transform to obtain the frequency-domain matching result.

[0018] In one embodiment, the method for identifying unauthorized charging of electric vehicles further includes:

[0019] Determine the first power consumption stage matching the charging stage from the total power consumption information according to the stage feature matching result;

[0020] Fuse the stage feature matching results of each stage to obtain the target matching result, including:

[0021] Determine the time intervals between the first power consumption stages;

[0022] Use the first power consumption stages corresponding to the time intervals within the preset time interval range as the associated power consumption stages;

[0023] Fuse the stage feature matching results corresponding to the associated power consumption stages of each stage to obtain the target matching result.

[0024] In one embodiment, perform identification of unauthorized charging of electric vehicles according to the target matching result, including:

[0025] Determine the second power consumption stage matching the charging stage from the total power consumption information according to the target matching result;

[0026] Determine the power consumption sequence and the total power consumption duration corresponding to each second power consumption stage among the second power consumption stages;

[0027] When the power consumption sequence is consistent with the charging sequence among the charging stages, and the total power consumption duration corresponding to each second power consumption stage is within the preset total charging duration range of the electric vehicle, it is determined that there is unauthorized charging of the electric vehicle at the user end.

[0028] Second aspect, the present application also provides an identification device for unauthorized connection of electric vehicles to chargers, including:

[0029] An acquisition module, configured to acquire the total electricity consumption information of the user side in the power system within a preset period;

[0030] A matching module, configured to respectively match the total electricity consumption information with multiple preset charging feature libraries for multiple different charging stages of electric vehicles to obtain stage feature matching results corresponding to each charging stage; fuse the multiple stage feature matching results to obtain a target matching result;

[0031] An identification module, configured to perform identification of unauthorized connection of electric vehicles to chargers according to the target matching result.

[0032] Third aspect, the present application also provides a computer device, including a memory and a processor, where the memory stores a computer program, and when the processor executes the computer program, the steps of the method for identifying unauthorized connection of electric vehicles to chargers in the first aspect are implemented.

[0033] Fourth aspect, the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method for identifying unauthorized connection of electric vehicles to chargers in the first aspect are implemented.

[0034] Fifth aspect, the present application also provides a computer program product, including a computer program, and when the computer program is executed by a processor, the steps of the method for identifying unauthorized connection of electric vehicles to chargers in the first aspect are implemented.

[0035] For the above-mentioned method, device, computer device, computer-readable storage medium, and computer program product for identifying unauthorized connection of electric vehicles to chargers, the total electricity consumption information of the user side in the power system within a preset period is used as the basis for identifying unauthorized connection of electric vehicles to chargers. This does not require dedicated equipment to be set at the user side, realizing non-intrusive identification of unauthorized connection of electric vehicles to chargers, which helps to improve the identification efficiency and reduce the identification cost; at the same time, starting from the charging characteristics of electric vehicles, different charging stages are divided to match the total electricity consumption information with the preset charging feature libraries to obtain stage feature matching results corresponding to each charging stage, thereby obtaining a target matching result, which further helps to achieve more accurate identification of unauthorized connection of electric vehicles to chargers. Description of the Drawings

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required to be used in the description of the embodiments of the present application or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0037] Figure 1 It is a schematic flow chart of a method for identifying unauthorized charging of electric vehicles in an embodiment;

[0038] Figure 2 It is another schematic flow chart of a method for identifying unauthorized charging of electric vehicles in an embodiment;

[0039] Figure 3 It is yet another schematic flow chart of a method for identifying unauthorized charging of electric vehicles in an embodiment;

[0040] Figure 4 It is a charging curve graph of an electric vehicle in an embodiment;

[0041] Figure 5 It is a charging curve graph corresponding to the first charging stage of an electric vehicle in an embodiment;

[0042] Figure 6 It is a charging curve graph corresponding to the second charging stage of an electric vehicle in an embodiment;

[0043] Figure 7 It is a charging curve graph corresponding to the third charging stage of an electric vehicle in an embodiment;

[0044] Figure 8 It is a structural block diagram of a device for identifying unauthorized charging of electric vehicles in an embodiment;

[0045] Figure 9 It is an internal structure diagram of a computer device in an embodiment. Detailed implementation manners

[0046] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application. Unless otherwise specified, the multiple referred to herein may represent two or more.

[0047] With the rapid popularization of electric vehicles, the proportion of electric vehicle charging behavior in household electricity consumption has increased significantly. However, electric vehicle charging behavior is characterized by high power, randomness, and long-term fluctuations. Its disorderly access may lead to problems such as local grid overload and three-phase imbalance, posing a severe challenge to the safe and stable operation of the distribution network. Especially for residential users, the behavior of privately connecting electric vehicle charging equipment often lacks effective supervision, exacerbating the peak-valley difference and line loss of the power grid. Although the traditional intrusive load monitoring technology (ILM) can accurately obtain device power consumption data through independent sensors, its deployment cost is high, maintenance is complex, and there is a risk of infringing on user privacy, making it difficult to be widely promoted on the residential side. In this context, non-intrusive load monitoring technology (NILM) only needs to collect the total power consumption signals (voltage, current, power, etc.) at the meter entrance, and can then decompose and identify the characteristics of each electrical load through algorithms, becoming the mainstream mode of load detection.

[0048] The first existing technical solution mainly uses the improved independent component analysis algorithm (ICA) to monitor the charging behavior of household electric vehicles. This method first analyzes the NILM system framework and the load characteristics of residential electric vehicles, and uses a method combining thresholding and filters to extract the load waveform during the stable charging process of electric vehicles. Then, on the basis of the traditional ICA algorithm, a modified third-order convergent Newton iteration algorithm is introduced to improve the update and iteration process of the ICA algorithm, making its convergence speed faster. Finally, the improved ICA algorithm is applied to the extraction process of non-intrusive household electric vehicle charging behavior to extract three stages of the charging behavior, thereby discriminating the electric vehicle charging behavior. The advantage of this solution is that by improving the ICA algorithm, the identification accuracy and speed of electric vehicle charging behavior are improved, but it still needs to be improved in dealing with the power overlap problem when multiple electrical appliances are running simultaneously.

[0049] The second prior art solution mainly utilizes the low-frequency characteristics of the electric vehicle charging load, adopts a two-stage decomposition technique to extract the low-frequency component of the smart meter, and combines event monitoring and the Dynamic Time Warping (DTW) method to estimate the charging time and amplitude. First, the STL (Seasonal and Trend decomposition using LOESS) method is used to extract the low-frequency trend component in the total load signal of the smart meter, and then the discrete wavelet transform (DWT) is further used to extract the feature component related to the electric vehicle charging load. Then, the high-order power difference is used to detect the edge of the charging event, determine the start and end times of charging, and match the reference template and the detection template through the DTW algorithm to determine the charging power. Finally, taking the decomposed electric vehicle charging load and the aggregated smart meter data as inputs, the CNN-Attention-LSTM neural network algorithm is used for training to predict the electric vehicle charging situation of household users in the short term. The advantage of this solution is that it makes full use of the low-frequency characteristics of the electric vehicle charging load and improves the identification accuracy, but it still has certain limitations in dealing with high-frequency noise and multi-appliance power overlap.

[0050] The third prior art solution mainly uses the steady-state and transient power characteristics of household appliances during normal operation as the reference quantity for closeness comparison, and designs a non-intrusive electric vehicle charge and discharge identification method. First, the Euclidean closeness is used to calculate the similarity between the load waveform and the template, and then the maximum membership principle is combined to determine the load type. Finally, by analyzing the charge and discharge behavior characteristics of the electric vehicle and using the conventional transient and steady-state parameters of the electric vehicle charge and discharge as the basic template, the charge and discharge state of the electric vehicle is identified. The advantage of this solution is that it can relatively quickly and accurately identify the charge and discharge behavior of the electric vehicle through closeness calculation and the maximum membership principle, but its identification accuracy in dealing with complex power consumption scenarios still needs to be improved.

[0051] Although existing non-intrusive load monitoring technologies have made some progress in identifying electric vehicle charging behaviors, there are still deficiencies such as insufficient dynamic feature modeling, lack of time-frequency domain feature fusion, and neglect of time continuity logic. Specifically: First, insufficient dynamic feature modeling: Existing methods (such as Existing Technical Solution 1 and Existing Technical Solution 3) mainly rely on static feature templates such as steady-state or single-stage charging waveforms, unable to capture the composite characteristics of periodic pulses and gradual change patterns, and failing to effectively model the multi-stage dynamic switching characteristics of electric vehicle charging (such as the periodic pulse and gradual change process of fast charging → constant voltage → trickle charging). This results in the charging power of electric vehicles being easily confused with the steady-state waveforms of high-power devices such as air conditioners and electric water heaters in complex electricity consumption scenarios, leading to misjudgments. Second, lack of time-frequency domain feature fusion: Existing solutions mostly adopt single-domain features (such as time-domain ICA decomposition in Existing Technical Solution 1 or low-frequency component analysis in Existing Technical Solution 2), lacking the joint modeling of time-domain periodic pulses and frequency-domain harmonic attenuation laws. For example, Technical Solution 2 extracts electric vehicle charging characteristics through low-frequency components but does not utilize the spectral characteristics of high-frequency pulses (such as fundamental frequency and harmonic distribution), resulting in insufficient robustness in high-noise or power overlap scenarios. Third, neglect of time continuity logic: The charging behavior of electric vehicles has strict temporal correlation (such as the fast charging stage must precede the constant voltage stage). However, existing technologies (such as proximity matching in Existing Technical Solution 3) only focus on the similarity of single-point features, lacking the joint analysis of multi-stage logical relationships within a time window, not modeling the temporal logic of multi-stage continuous occurrences, and unable to distinguish accidental power fluctuations from real charging events.

[0052] The above deficiencies lead to low identification accuracy of electric vehicle charging behaviors in complex electricity consumption scenarios with overlapping multi-appliance powers in the existing technology. Therefore, there is an urgent need for a lightweight and highly robust NILM model that can accurately identify the hidden behavior of users' unauthorized connection of electric vehicle charging through multi-dimensional feature fusion and dynamic behavior modeling.

[0053] Based on the above analysis, this application provides a method for identifying unauthorized electric vehicle charging, which will be described below by way of embodiments:

[0054] In one embodiment, as Figure 1 shown, a method for identifying unauthorized electric vehicle charging is provided. In this embodiment, taking the application of this method to a server as an example, it can be understood that this method can also be applied to a system including a terminal and a server. Of course, this method can also be applied to power systems, power dispatching systems, systems or platforms related to power operation management, etc. In this embodiment, the method includes the following steps:

[0055] Step S101, the server obtains the total electricity consumption information of the user side in the power system within a preset period.

[0056] Among them, the user side can be the end that consumes electricity in the power system, which can be compared with the power generation side. Exemplarily, the total electricity meter corresponding to a household can be regarded as a user side.

[0057] Among them, the preset period can be a period customized according to actual needs. For example, one day, one week, etc. can be taken as a period.

[0058] Among them, the total electricity consumption information can be the total information about the electricity usage of the user side within the preset period. In some embodiments, since non-invasive identification technology is used for load identification, the obtained electricity-related information can be the total electricity consumption information of a user side, without the need to obtain detailed electricity usage information of each electrical device within the user side, etc. Exemplarily, the total electricity consumption information can include the electricity usage information of various household electrical appliances such as air conditioners and refrigerators at the user side, as well as the electricity usage information corresponding to privately connecting an electric vehicle for charging, etc.

[0059] Step S102, the server matches the total electricity consumption information with multiple charging feature libraries preset for multiple different charging stages of the electric vehicle respectively, and obtains the stage feature matching results corresponding to each charging stage.

[0060] Among them, the electric vehicle can be a vehicle with a charging requirement, including but not limited to pure electric vehicles and plug-in hybrid vehicles, etc.

[0061] Among them, the charging stage is for the charging of the electric vehicle. Exemplarily, the charging stage of the electric vehicle charging can be divided into multiple charging stages, such as, slow charging stage - fast charging stage - slow charging stage. In some embodiments, the charging stages corresponding to different types of electric vehicles can be different or the same.

[0062] Among them, the charging feature library can be preset for the charging stage. For example, if the charging stage includes charging stages 1, 2, 3, and 4, then the corresponding charging feature libraries can include charging feature libraries 1, 2, 3, and 4. In some embodiments, the charging feature library can be obtained by extracting features based on the power information corresponding to the electric vehicle during charging. In some embodiments, multiple charging feature libraries can be combined to form a total charging feature library. In some embodiments, different types or models of electric vehicles can correspond to different total charging feature libraries.

[0063] In some embodiments, the server may match the total electricity consumption information with multiple pre-set charging feature libraries for multiple different charging stages of each type of electric vehicle respectively, to obtain the stage feature matching results corresponding to each charging stage of each type of electric vehicle. And, based on the matching degrees of the stage feature matching results corresponding to each type of electric vehicle, determine the stage feature matching result corresponding to the target type of electric vehicle. For example, take the type of electric vehicle corresponding to the stage feature matching result with the highest matching degree as the target type of electric vehicle.

[0064] In some embodiments, the server may perform the matching between the total electricity consumption information and multiple charging feature libraries based on an artificial intelligence model, so as to more quickly determine multiple stage feature matching results.

[0065] Step S103, the server fuses multiple stage feature matching results to obtain a target matching result.

[0066] In some embodiments, the server may splice multiple stage feature matching results to obtain a target matching result.

[0067] In some embodiments, the server may determine whether there are repetitions, overlaps, etc. among multiple stage feature matching results, merge the stage feature matching results with repetitions and overlaps, and then integrate the merged stage feature matching results and the unmerged stage feature matching results to obtain a target matching result.

[0068] Step S104, the server performs identification of unauthorized charging of electric vehicles according to the target matching result.

[0069] In some embodiments, the server may determine the similarity between the total electricity consumption information and the charging information generated by unauthorized charging of electric vehicles according to the target matching result. If the similarity reaches a pre-set similarity range, such as greater than 75%, it is considered that there is unauthorized charging of electric vehicles; otherwise, there is no unauthorized charging of electric vehicles.

[0070] In some embodiments, the server may repeat the foregoing steps S101 to S103 to determine multiple target matching results, and comprehensively determine whether there is unauthorized charging of electric vehicles according to the multiple target matching results.

[0071] The above technical solution uses the total power consumption information of the user side in the power system within a preset period as the basis for identifying unauthorized electric vehicle charging. This does not require dedicated equipment to be set up at the user side, achieving non-intrusive identification of unauthorized electric vehicle charging, which helps improve the identification efficiency and reduce the identification cost. At the same time, starting from the charging characteristics of electric vehicles, different charging stages are divided to match the total power consumption information with a preset charging feature library, obtaining the stage feature matching results corresponding to each charging stage, thereby obtaining the target matching result, which further helps to achieve more accurate identification of unauthorized electric vehicle charging.

[0072] In one embodiment, the aforementioned total power consumption information may include power information; the aforementioned charging feature library may include the time domain features and frequency domain features of the charging power corresponding to each charging stage; the aforementioned "matching the total power consumption information with a preset charging feature library for different charging stages of electric vehicles to obtain the stage feature matching results corresponding to each charging stage" may include: for the charging feature library corresponding to each charging stage, matching the power information with the time domain features to obtain the time domain matching result; matching the power information with the frequency domain features to obtain the frequency domain matching result; and determining the stage feature matching result corresponding to the charging stage according to the time domain matching result and the frequency domain matching result.

[0073] Among them, the power information may be information characterizing the power consumption of the user side. Exemplarily, the power information may include the change information of power over time.

[0074] Among them, the time domain features may be features characterizing the change of the charging power corresponding to the charging stage of the electric vehicle over time. For example, the time features may include the maximum power value and the corresponding time point, the minimum power value and the corresponding time point, etc.

[0075] Among them, the frequency domain features may be features characterizing the charging power corresponding to the charging stage of the electric vehicle in terms of frequency.

[0076] In some embodiments, when the time domain matching degree characterized by the time domain matching result reaches the preset time domain matching degree, and the frequency domain matching degree characterized by the frequency domain matching result reaches the preset frequency domain matching degree, the server combines the time domain matching result and the frequency domain matching result to determine the stage feature matching result corresponding to the charging stage. If the time domain matching degree does not reach the preset time domain matching degree, and / or the frequency domain matching degree does not reach the preset frequency domain matching degree, the time domain matching result and the frequency domain matching result are re-determined. This helps to determine a more accurate stage feature matching result.

[0077] In some embodiments, "determining the stage feature matching result corresponding to the charging stage according to the time domain matching result and the frequency domain matching result" may include:

[0078] When the time-domain matching degree characterized by the time-domain matching result reaches a preset time-domain matching degree, it is determined that the time-domain feature matching is successful, and the time period corresponding to the matched time-domain feature is recorded as the time-domain time period; and,

[0079] When the frequency-domain matching degree characterized by the frequency-domain matching result reaches a preset frequency-domain matching degree, it is determined that the frequency-domain feature matching is successful, and the time period corresponding to the matched frequency-domain feature is recorded as the frequency-domain time period; and,

[0080] When the time-domain time period is consistent with the frequency-domain time period, according to the time-domain matching result and the frequency-domain matching result, the stage feature matching result corresponding to the charging stage is determined.

[0081] Starting from both the time domain and the frequency domain, the total electricity consumption information is matched with the characteristics of each charging stage of the electric vehicle to obtain the time-domain matching result and the frequency-domain matching result, so as to determine the stage feature matching result, and thus a more accurate stage feature matching result can be obtained. This can achieve composite feature matching with stronger anti-interference ability and reduce the misjudgment rate in the scenario of multi-electrical appliance power overlap.

[0082] In one embodiment, the aforementioned "matching the power information with the time-domain feature to obtain the time-domain matching result" may include: the server determines the pulse period, duty cycle, and pulse amplitude in the power information; according to the pulse period, duty cycle, and pulse amplitude, the power information is matched with the time-domain feature to obtain the time-domain matching result.

[0083] Among them, the pulse period refers to the time interval between two adjacent pulses in a periodically repeating pulse sequence. For example, the pulse period can include two parts: high level and low level, that is, the time required for the pulse signal to complete a full fluctuation.

[0084] Among them, the duty cycle refers to the ratio of the duration of the effective state (such as high level) to the entire cycle time in a periodic signal.

[0085] Among them, the pulse amplitude refers to the difference between the maximum value and the minimum value of the signal in a pulse signal.

[0086] In some embodiments, based on the power information, a pulse signal can be determined, so as to determine the corresponding pulse period, duty cycle, and pulse amplitude.

[0087] By matching the power information with the time-domain feature according to the pulse period, duty cycle, and pulse amplitude, the server can obtain a more accurate time-domain matching result.

[0088] In one embodiment, the foregoing "matching the power information with the frequency-domain features to obtain a frequency-domain matching result" may include: The server matches the power information with the frequency-domain features according to the short-time Fourier transform to obtain a frequency-domain matching result.

[0089] Among them, the short-time Fourier transform (STFT, short-time Fourier transform, or short-term Fourier transform) is a mathematical transform related to the Fourier transform, used to determine the frequency and phase of the sine wave in the local region of a time-varying signal.

[0090] By using the short-time Fourier transform, the server can match the power information with the frequency-domain features to obtain a more accurate frequency-domain matching result.

[0091] In one embodiment, the foregoing method for identifying unauthorized charging of electric vehicles may further include: The server determines a first power consumption stage that matches the charging stage from the total power consumption information according to the stage feature matching result; the foregoing "fusing the stage feature matching results to obtain a target matching result" may include: The server determines the time intervals between the first power consumption stages; takes the first power consumption stages corresponding to the time intervals within a preset time range as associated power consumption stages; and fuses the stage feature matching results corresponding to the associated power consumption stages to obtain a target matching result.

[0092] In some embodiments, the first power consumption stage correspondingly includes each stage during the charging process of the electric vehicle, such as stage one, stage two, and stage three.

[0093] In some embodiments, in the process of matching the total power consumption information with each charging stage of the electric vehicle, similar to the fact that there can be multiple charging stages for the electric vehicle charging, the total power consumption information can also correspondingly have multiple power consumption stages, so that the power consumption stages can be matched with the charging. That is, the server determines a first power consumption stage that matches the charging stage from the total power consumption information according to the stage feature matching result.

[0094] In some embodiments, the multiple first power consumption stages may be continuous, discontinuous, or partially continuous.

[0095] In some embodiments, during the actual charging process of the electric vehicle, situations such as continuing to charge after a short stop, the transition of the charging stage, or a short interruption may occur. Therefore, the server can determine the time intervals between the first power consumption stages; if the time interval is within a preset time range, such as 5 minutes, it indicates that there may be a situation of continuing to charge after a short stop. In this regard, the corresponding first power consumption stage can be taken as an associated power consumption stage, and the stage feature matching results corresponding to the associated power consumption stages are fused to obtain a target matching result.

[0096] The server determines a first power consumption stage that matches the charging stage from the total power consumption information, determines the time intervals between the first power consumption stages, and uses the first power consumption stages corresponding to the time intervals within a preset time range as associated power consumption stages; fuses the stage feature matching results corresponding to the associated power consumption stages to obtain a target matching result. This can handle situations such as transitions or short interruptions during charging stage switching and avoid affecting the accuracy of the target matching result due to the above situations.

[0097] In one embodiment, as Figure 2 shown, the aforementioned "identifying unauthorized electric vehicle charging based on the target matching result" may include steps S201 to S203:

[0098] Step S201: The server determines a second power consumption stage that matches the charging stage from the total power consumption information according to the target matching result.

[0099] In some embodiments, the second power consumption stage may correspond to each stage during the charging process of the electric vehicle, such as stage one, two, and three. In some embodiments, the second power consumption stage may be determined based on the first power consumption stage.

[0100] In some embodiments, similar to the determination of the first power consumption stage based on the stage feature matching result, the second power consumption stage can be determined according to the target matching result. Since the target matching result is more accurate than the stage feature matching result, the determined second power consumption stage can better match the charging stage.

[0101] Step S202: The server determines the power consumption sequence and the total power consumption duration corresponding to each second power consumption stage among the second power consumption stages.

[0102] In some embodiments, among the multiple second power consumption stages, they may be continuous, discontinuous, or partially continuous.

[0103] Step S203: When the power consumption sequence is consistent with the charging sequence between the charging stages and the total power consumption duration corresponding to each second power consumption stage is within the preset total charging duration range of the electric vehicle, the server determines that there is unauthorized electric vehicle charging at the user end.

[0104] Among them, the charging sequence between the charging stages can be preset or obtained in advance. In some embodiments, different types of electric vehicles may correspond to different charging sequences between the charging stages.

[0105] In some embodiments, if the sequence of electricity consumption is inconsistent with the sequence of charging in each charging stage, and / or the total duration of electricity consumption corresponding to each second electricity consumption stage is not within the preset total duration range of electric vehicle charging, it is determined that there is no unauthorized connection of an electric vehicle at the user side.

[0106] The server determines the second electricity consumption stage according to the target matching result to verify the consistency between the sequence of electricity consumption and the sequence of charging, and verify whether the total duration of electricity consumption is within the preset total duration range of electric vehicle charging. By comprehensively considering these two aspects, namely the sequence of electricity consumption and the total duration of electricity consumption, it is determined whether there is unauthorized connection of an electric vehicle at the user side, which helps to improve the accuracy of identifying unauthorized connection of an electric vehicle.

[0107] In an exemplary embodiment, a method for identifying unauthorized connection of an electric vehicle is provided. This method can correspond to a non-intrusive load monitoring model for identifying the unauthorized connection of an electric vehicle within a user, that is, this method can be implemented based on this model. The purpose of this method and this model is to achieve the following goals: First, multi-stage dynamic feature modeling: By analyzing the periodic pulse waveforms and spectral characteristics of multiple charging stages, the problem of feature confusion caused by power overlap of multiple electrical appliances is solved. Second, time-domain and frequency-domain feature fusion: A time-domain and frequency-domain joint feature library for electric vehicle charging is constructed, and the time-domain pulse period and frequency-domain harmonic attenuation law are jointly analyzed to improve the robustness in the power overlap scenario and enhance the anti-noise ability. Third, timing correlation analysis: A timing clustering algorithm that tolerates redundancy is designed to verify the logical correlation within a time window for the detected multi-stage features, so as to distinguish real charging events from random power fluctuations. Through the double innovation of multi-dimensional feature modeling and timing logic constraints, this method solves the core problems of the existing NILM technology in identifying unauthorized connection of an electric vehicle, such as the lack of dynamic features and the neglect of timing correlation, and provides a low-cost and high-robustness load monitoring solution for the grid side. Specifically:

[0108] I. Overall process structure:

[0109] Task 1: Mathematical modeling of the feature sets in each stage. It is necessary to construct an electric vehicle charging feature library (corresponding to the aforementioned preset multiple charging feature libraries), the purpose of which is to provide a comparison benchmark for detection. Based on the existing electric vehicle charging waveform data, the charging waveform is cut by stage, and independent segments of stage one, stage two, and stage three are extracted respectively. Then, time-domain analysis and frequency-domain analysis are performed on each stage respectively, time-domain features and frequency-domain features are extracted, and the time-domain and frequency-domain feature parameters of each stage are integrated into a structured feature library. Through this step of design, a multi-stage dynamic characteristic library (corresponding to the aforementioned preset multiple charging feature libraries) is constructed, and by combining time-domain and frequency-domain features, the identification accuracy of the model is improved.

[0110] Task 2: Feature matching in the mixed waveform. Then design the electric vehicle charging detection process, aiming to identify behaviors that conform to the electric vehicle charging characteristics from the user's total daily electricity consumption data. Before the detection starts, the original total power data will be denoised to eliminate the influence of grid noise on feature extraction. After data preprocessing, based on the constructed electric vehicle charging feature library mentioned above, feature matching will be carried out. Only when both the time-domain and frequency-domain features match, will the window be marked as belonging to a certain charging stage, and thus record all time periods that conform to Stages 1, 2, and 3.

[0111] Task 3: Clustering judgment of consecutive stages. For the time periods marked as Stages 1, 2, and 3, perform clustering and merging in chronological order to verify whether they conform to the timing logic of "Stage 1 → Stage 2 → Stage 3". And a redundant design is made, that is, a certain period of time interval or overlap between each stage is allowed. Such a design takes into account that in real charging events, there may be a transition interval during stage switching, and strict sequential matching will lead to missed detections.

[0112] After completing the above process, all verified charging events can be output, including the start and end times and the duration of each stage. The overall process structure is as follows Figure 3 shown.

[0113] II. Detailed modeling method:

[0114] First of all, this model is constructed based on the charging curve of a known electric vehicle model, and this charging curve comes from a public dataset, which consists of the maximum power, the minimum power, and the average actual power measured every 1 minute. As follows Figure 4 shown:

[0115] Task 1: Mathematical modeling of the feature sets for each stage

[0116] 1. Stage division and definition of piecewise function

[0117] Referring to Figure 4 , a possible charging curve corresponding to an electric vehicle is given. This model divides the electric vehicle charging into three stages, and the power function of each stage can be modeled as a piecewise function, as shown in formula (1):

[0118]

[0119] In the formula, P(t) is the charging power signal, and the total charging duration is T = T1 + T2 + T 3， Among them:

[0120]

[0121] 2. Feature set of charging stage 1

[0122] 2.1 Modeling of time-domain features in charging stage 1

[0123] The first charging stage is characterized by high-frequency pulse cycles, as can be seen in Figure 5 shown. The charging power in the first charging stage is as shown in Equation (2):

[0124]

[0125] where the pulse period τ1 = Thigh + Tlow, Thigh is the high-level duration, approximately 5 minutes, Tlow is the low-level transition time, approximately 10 seconds, and the pulse amplitude is P max (maximum power), P min (valley power), and N is the number of pulses.

[0126] The duty cycle D1 in the first charging stage is as shown in Equation (3):

[0127]

[0128] The number of pulses N in the first charging stage is as shown in Equation (4):

[0129]

[0130] 2.2 Frequency-domain feature modeling of the first charging stage

[0131] By analyzing the spectral characteristics of periodic pulses through Fourier transform, the fundamental frequency in this stage is as shown in Equation (5):

[0132]

[0133] The amplitude of its harmonic component, i.e., the nth harmonic, is as shown in Equation (6):

[0134] An = (Pmax - Pmin) × sinc(nD1) (6)

[0135] Then, the frequency-domain feature set of the first charging stage can be obtained, as shown in Equation (7):

[0136]

[0137] where δ(f - nf1) is the Dirac δ function.

[0138] 3. Feature set of the second charging stage

[0139] 3.1 Time-domain feature modeling of the second charging stage

[0140] The second charging stage is characterized by low-frequency pulse cycles, as can be seen in Figure 6 shown.

[0141] Then, the charging power in the second charging stage is as shown in Equation (8):

[0142]

[0143] Wherein, the pulse period τ2 = T’high + T’low, T’high is the high-level duration, about 15 minutes, T’low is the low-level transition time, about 10 seconds, and the pulse amplitude is P max (maximum power), P min (valley power), and M is the number of pulses.

[0144] The duty cycle D2 of the second charging stage is as shown in formula (9):

[0145]

[0146] The number of pulses M in the second charging stage is as shown in formula (10):

[0147]

[0148] 3.2 Frequency-domain feature modeling of the second charging stage

[0149] By analyzing the spectrum characteristics of the periodic pulse through Fourier transform, the fundamental frequency of this stage is as shown in formula (11):

[0150]

[0151] The amplitude of its harmonic component, i.e., the nth harmonic, is as shown in formula (12):

[0152] A' n =(P max -P min ) × sinc(nD2) (12)

[0153] Then the frequency-domain feature set of the second charging stage can be obtained, as shown in formula (13):

[0154]

[0155] 4. Feature set of the third charging stage

[0156] 4.1 Time-domain feature modeling of the third charging stage

[0157] The third charging stage shows a linear decline and an end cyclic pulse, as can be seen in Figure 7 as shown.

[0158] The charging power in the third charging stage is as shown in formula (14):

[0159]

[0160] Wherein, the time T of the first half linear decline segmentdown It is about 5 minutes, and when it drops to 1 / 4 of the maximum power, it switches to pulse cycling. The pulse period τ3 = T”high + T”low, where T”high is the high-level duration, about 5 minutes, and T”low is the low-level transition time, about 10 seconds. The pulse amplitude is P max / 4 (1 / 4 of the maximum power), P min (valley power), and K is the number of pulses. After K pulses end, it drops to the valley power, indicating the end of charging.

[0161] The duty cycle D3 of the pulse cycling in the third charging stage is as shown in formula (15):

[0162]

[0163] The number of pulses K of the pulse cycling in the third charging stage is as shown in formula (16):

[0164]

[0165] 4.3 Modeling of the frequency-domain characteristics in the third charging stage

[0166] By analyzing the spectral characteristics of the periodic pulse through Fourier transform, the fundamental frequency in this stage is as shown in formula (17):

[0167]

[0168] The amplitude of its harmonic component, that is, the nth harmonic, is as shown in formula (18):

[0169]

[0170] Then the frequency-domain characteristic set in the third charging stage can be obtained, as shown in formula (19):

[0171]

[0172] where F down (f) is the spectral contribution of the linearly decreasing segment, mainly containing low-frequency components. The spectrum of the last cycle shows a single pulse (without a rising signal), and the harmonic energy is significantly reduced, indicating the end of charging.

[0173] Task 2: Feature matching in the mixed waveform

[0174] Through Task 1, a time-frequency comprehensive feature library for electric vehicle charging is constructed (corresponding to the previously preset multiple charging feature libraries). Task 2 will focus on using this feature library to perform feature matching on the user's total daily power data. First, the user's total daily power data should be processed by moving average filtering and normalization to eliminate high-frequency noise and dimensional differences. This data preprocessing process is not the focus of this application, so it will not be elaborated here.

[0175] 1. Time-domain Feature Matching Algorithm

[0176] This application uses the sliding window matching method for time-domain feature matching. First, the all-day power sequence P(t) and the window length Win are input, and then the following metrics are calculated for each window in P(t):

[0177] Pulse period detection: Search for periodic peaks through the autocorrelation function.

[0178] Duty cycle calculation: Statistically calculate the proportion of high-power time.

[0179] Amplitude matching: Verify whether Pmax and Pmin meet the expectations.

[0180] Then the matching condition is shown in formula (20):

[0181]

[0182] Among them, ρ(Δt) is the autocorrelation function used to detect the period τ, and ∈ is the matching error threshold.

[0183] If the metrics within the window match a certain stage feature set, mark that time period.

[0184] This step corresponds to the step of determining the time-domain matching result described above.

[0185] 2. Frequency-domain Feature Matching Algorithm

[0186] This application uses the short-time Fourier transform (STFT) for time-domain feature matching:

[0187] Frame P(t), and set the frame length to 10 minutes.

[0188] Calculate the power spectral density for each frame

[0189] Detect whether the fundamental frequencies f1, f2, f3 exist and verify the harmonic attenuation mode

[0190] If the spectral features match a certain stage, record the corresponding time points.

[0191] This step corresponds to the step of determining the frequency-domain matching result described above.

[0192] 3. Joint Feature Matching

[0193] Time-domain + frequency-domain joint matching: Only when both match a certain stage, it is determined that the matching for that stage is successful. Thus, the stage feature matching result described above is obtained.

[0194] Output: Record all time periods {S1, S2, S3} that meet the characteristics of Phase 1, Phase 2, and Phase 3. For each time period Si, record the specific start and end times {[ti,start, ti,end],...}.

[0195] Task 3: Clustering judgment for consecutive phases

[0196] From Task 2, all time periods {S1, S2, S3} that meet the characteristics of Phase 1, Phase 2, and Phase 3 are obtained. Then, the interval between the end time ti,end of Phase i and the start time tj,start of Phase j (corresponding to the time interval between the first power consumption phases mentioned above) is shown in formula (21):

[0197] d(Si, Sj) = |ti,end - tj,start| (21)

[0198] At this time, the merging condition is that if S i is Phase 1, S j is Phase 2, and formula (22) is satisfied, then S i and S j are merged into a candidate event:

[0199] d(Si, Sj) ≤ Δttol (22)

[0200] where Δttol is the tolerance threshold (corresponding to the preset time interval range mentioned above), that is, it allows for the redundant design of the end time of S i and the start time of S j to have an interval or overlap, adapting to the transition or short interruption of the charging phase switching in the real scenario and preventing missed detections. Iterative merging is performed according to the above rules until all possible mergings are completed.

[0201] Result output:

[0202] Through the clustering judgment of Task 3, the result output only retains the clustering clusters that meet the order of "Phase 1 → Phase 2 → Phase 3" (corresponding to the power consumption sequence and the charging sequence between each charging phase being consistent as mentioned above), and at the same time, the total duration of this cluster meets the expected range (such as 2 - 4 hours) (corresponding to the total power consumption duration of the second power consumption phase being within the preset total charging duration range of the electric vehicle), then it is confirmed as an electric vehicle charging event, and finally, a charging event list is output, including start and end times, phase distribution, etc.

[0203] The above technical solution has the following effects: 1. Multi-stage dynamic feature modeling: By extracting the dynamic waveform features of electric vehicle charging in stages, it solves the problem of insufficient modeling of complex charging behaviors by traditional methods. Specifically, the existing technology relies on single-stage or steady-state feature templates and cannot distinguish the similar waveforms of electric vehicle charging and high-power devices such as air conditioners. However, through the dynamic characteristics extracted in stages in this application, combined with the time-domain duty cycle and frequency-domain harmonic attenuation law, the feature discrimination degree is significantly improved. For example, although the power fluctuation of the air conditioner is similar to that of fast charging, it lacks periodic pulses and specific harmonic attenuation patterns and can be excluded by the combined features. 2. Time-domain - frequency-domain feature fusion: By combining the time-domain pulse characteristics (period, duty cycle) with the frequency-domain harmonic attenuation law (fundamental frequency, harmonic distribution), it realizes a composite feature matching with stronger anti-interference ability and significantly reduces the misjudgment rate in the scenario of multi-electrical appliance power overlap. Specifically, the existing technology relies on low-frequency components and deep learning models, is sensitive to high-frequency noise (such as the transient of switched electrical appliances) and has complex calculations. Through the dual constraints of time-domain pulse period detection + frequency-domain harmonic energy verification in this application, it not only avoids high-frequency noise interference (frequency-domain filtering) but also makes up for the limitations of single-domain features. For example, in the power overlap scenario, the short-term peak of the microwave oven may falsely trigger the time-domain detection, but its wide-frequency spectrum characteristic is significantly different from the harmonic law of electric vehicle charging and can be filtered by the frequency-domain rule. 3. Redundant time-series clustering algorithm: Design a time-series logic verification mechanism that allows stage intervals or partial overlaps to ensure the continuity of real charging events and improve the adaptability of the model to actual scenarios.

[0204] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are shown in sequence according to the indications of the arrows, these steps do not necessarily have to be executed in the order indicated by the arrows. Unless there is a clear indication in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages do not necessarily have to be executed at the same time, but can be executed at different times. The execution order of these steps or stages does not necessarily have to be sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0205] Based on the same inventive concept, the embodiments of this application also provide an electric vehicle unauthorized charging recognition device for implementing the above-mentioned electric vehicle unauthorized charging recognition method. The solution provided by this device to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the electric vehicle unauthorized charging recognition device provided below can refer to the limitations on the electric vehicle unauthorized charging recognition method in the above text and will not be repeated here.

[0206] In an exemplary embodiment, as Figure 8 shown, an electric vehicle unauthorized charging identification device 800 is provided, including:

[0207] An acquisition module 801, configured to acquire the total power consumption information of the user side in the power system within a preset period;

[0208] A matching module 802, configured to respectively match the total power consumption information with a plurality of pre-set charging feature libraries for multiple different charging stages of the electric vehicle to obtain stage feature matching results corresponding to each charging stage; and fuse the plurality of stage feature matching results to obtain a target matching result;

[0209] An identification module 803, configured to perform electric vehicle unauthorized charging identification according to the target matching result.

[0210] In one embodiment, the foregoing total power consumption information includes power information; the charging feature libraries include time domain features and frequency domain features of the charging power corresponding to each charging stage;

[0211] The matching module 802 is further configured to match the total power consumption information with a pre-set charging feature library for different charging stages of the electric vehicle to obtain stage feature matching results corresponding to each charging stage, including: for the charging feature library corresponding to each charging stage, matching the power information with the time domain features to obtain a time domain matching result; matching the power information with the frequency domain features to obtain a frequency domain matching result; and determining the stage feature matching result corresponding to the charging stage according to the time domain matching result and the frequency domain matching result.

[0212] In one embodiment, the matching module 802 is further configured to match the power information with the time domain features to obtain a time domain matching result, including: determining the pulse period, duty cycle, and pulse amplitude in the power information; and matching the power information with the time domain features according to the pulse period, duty cycle, and pulse amplitude to obtain a time domain matching result.

[0213] In one embodiment, the matching module 802 is further configured to match the power information with the frequency domain features to obtain a frequency domain matching result, including: matching the power information with the frequency domain features according to the short-time Fourier transform to obtain a frequency domain matching result.

[0214] In one embodiment, the matching module 802 is further configured to determine, from the total power consumption information, a first power consumption stage that matches the charging stage according to the stage feature matching result; fuse the stage feature matching results to obtain a target matching result, including: determining the time intervals between the first power consumption stages; taking the first power consumption stages corresponding to the time intervals within a preset time interval range as associated power consumption stages; and fusing the stage feature matching results corresponding to the associated power consumption stages to obtain the target matching result.

[0215] In one embodiment, the identification module 803 is further configured to perform identification of unauthorized electric vehicle charging according to the target matching result, including: determining, from the total power consumption information, a second power consumption stage that matches the charging stage according to the target matching result; determining the order of power consumption among the second power consumption stages and the total power consumption duration corresponding to each second power consumption stage; and determining that there is unauthorized electric vehicle charging at the user side when the order of power consumption is consistent with the charging order among the charging stages and the total power consumption duration corresponding to each second power consumption stage is within the preset total electric vehicle charging duration range.

[0216] Each module in the above unauthorized electric vehicle charging identification device can be implemented in whole or in part by software, hardware, and their combination. The above modules can be embedded in the processor in the computer device in the form of hardware or be independent of the processor, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0217] In an exemplary embodiment, a computer device is provided. The computer device can be a server, and its internal structure diagram can be as Figure 9 shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the data required for executing the unauthorized electric vehicle charging identification method, such as the total power consumption information, etc. The input / output interface of the computer device is used for exchanging information between the processor and external devices. The communication interface of the computer device is used for communicating with external terminals through a network connection. The computer program, when executed by the processor, implements an unauthorized electric vehicle charging identification method.

[0218] Those skilled in the art can understand thatFigure 9 The structure shown is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0219] In an exemplary embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0220] In an embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0221] In an embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0222] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include Read-Only Memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, Resistive RandomAccess Memory (ReRAM), MagnetoresistiveRandomAccess Memory (MRAM), Ferroelectric RandomAccess Memory (FRAM), Phase Change Memory (PCM), graphene memory, etc. Volatile memory can include Random Access Memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as Static RandomAccess Memory (SRAM) or Dynamic RandomAccess Memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logic devices, data processing logics based on quantum computing, Artificial Intelligence (AI) processors, etc., without limitation.

[0223] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered to be within the scope recorded in the present application.

[0224] The above-described embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. A method for identifying private charging of electric vehicles, characterized in that: The method comprises: Obtaining total power consumption information of the user end in the power system within a preset period; Matching the total power consumption information with a plurality of charging feature libraries preset for a plurality of different charging stages of the electric vehicle respectively, to obtain a stage feature matching result corresponding to each of the charging stages; The feature matching results of multiple stages are merged to obtain a target matching result; According to the target matching result, the electric vehicle private charging is identified.

2. The method according to claim 1, characterized in that The total power consumption information includes power information; the charging feature library includes the time domain characteristics and frequency domain characteristics of the charging power corresponding to each charging stage; The total power consumption information is matched with a charging feature library preset for different charging stages of the electric vehicle to obtain a stage feature matching result corresponding to each charging stage, including: For the charging feature library corresponding to each charging stage, the power information is matched with the time domain feature to obtain a time domain matching result; the power information is matched with the frequency domain feature to obtain a frequency domain matching result; The stage feature matching result corresponding to the charging stage is determined according to the time domain matching result and the frequency domain matching result.

3. The method according to claim 2, characterized in that The matching of the power information with the time domain feature to obtain a time domain matching result includes: Determine the pulse period, duty cycle and pulse amplitude in the power information; According to the pulse period, the duty cycle and the pulse amplitude, the power information is matched with the time domain feature to obtain a time domain matching result.

4. The method according to claim 2, characterized in that: The matching of the power information with the frequency domain feature to obtain a frequency domain matching result includes: According to short-time Fourier transform, the power information is matched with the frequency domain feature to obtain a frequency domain matching result.

5. The method according to claim 1, characterized in that: The method further comprises: According to the stage feature matching result, determining a first power consumption stage matching the charging stage from the total power consumption information; The step of fusing the feature matching results of each stage to obtain the target matching result includes: determining a time interval between each of the first power consumption stages; taking the first power consumption stage corresponding to the time interval within the preset time interval range as the associated power consumption stage; The stage feature matching results corresponding to each of the associated power consumption stages are fused to obtain a target matching result.

6. The method according to any one of claims 1 to 5, characterized in that: The step of identifying the private charging connection of the electric vehicle according to the target matching result includes: According to the target matching result, determining a second power usage stage matching the charging stage from the total power usage information; Determine the power usage sequence between the second power usage stages, and the total power usage duration corresponding to the second power usage stages; When the electricity usage sequence is consistent with the charging sequence between the charging stages, and the total electricity usage time corresponding to each of the second electricity usage stages is within a preset total electric vehicle charging time range, it is determined that the electric vehicle is privately charged at the user end.

7. A device for identifying private charging of electric vehicles, characterized in that: The device comprises: An acquisition module is used to acquire total power consumption information of a user end in a preset period in the power system; A matching module, used to match the total power consumption information with a plurality of charging feature libraries preset for a plurality of different charging stages of the electric vehicle, respectively, to obtain a stage feature matching result corresponding to each of the charging stages; and to fuse the plurality of stage feature matching results to obtain a target matching result; The identification module is used to identify the electric vehicle privately connected for charging according to the target matching result.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images