CN114759558A - Non-invasive online rapid detection method for charging load of electric bicycle - Google Patents

Abstract

The invention discloses a non-invasive online rapid detection method for a charging load of an electric bicycle, which comprises the following steps: and sequentially carrying out state conversion removal, down-sampling and SG filtering on the power data of the electric bicycle with the frequency of 1Hz for independent charging, respectively calculating the difference signals of active/reactive power, and constructing a charging load template of the electric bicycle by utilizing a fitting curve of the difference signals in a constant voltage stage. Acquiring active/reactive power data of a user at a home, taking a time window as 6 hours and a step length as 10 minutes, sequentially performing state conversion removal, down-sampling and SG filtering in the window, respectively calculating differential signals of active/reactive power, matching continuous negative subsequences in the active power differential signals with continuous positive subsequences in the reactive power differential signals, calculating the distance between the successfully matched subsequences and a template, and judging that the electric bicycle charging load exists at the position if the distance is smaller than a distance threshold value. The method has wide application prospect in the fields of illegal charging inspection of the electric bicycle and the like.

Description

On-line rapid detection method of non-intrusive electric bicycle charging load

technical field

The invention relates to the field of electric bicycle charging load monitoring, in particular to a non-invasive online fast detection method of electric bicycle charging load based on local features.

Background technique

Under the unified deployment of the "dual carbon" strategic goal, clean electric energy replacement has become an important way to get rid of fossil energy dependence. In the field of transportation, my country's new energy vehicle and electric bicycle industries have also entered a period of rapid development. Among them, the social ownership of electric bicycles has exceeded 300 million. However, due to the lack of planning and management of electric bicycle charging sites and the weak safety awareness of residents, related fire accidents frequently occur, often causing huge casualties and property losses. For this reason, power, property and other departments often need to manually detect users' illegal charging behaviors on the spot, but there are problems such as low efficiency and low user cooperation. Non-intrusive Load Monitoring (NILM) technology does not need to intrude into the user's interior. It only needs to process and analyze the total load power consumption data to obtain the detailed power consumption information of each electrical appliance of the user. This analyzes the user's electricity consumption behavior. Therefore, it is very practical to apply NILM technology to the efficient detection of illegal charging of electric bicycles. This efficient and convenient monitoring technology will also have broad applications in the fields of electric bicycle health status assessment, charging power query, and energy efficiency analysis. prospect.

At present, the application of non-intrusive load monitoring technology in residential users mainly focuses on some common household appliances, and there are few studies on electric bicycle charging load detection methods. And although the state-of-the-art NILM method can decompose most of the household appliance loads, the electric bicycle charging load is a continuously variable load, and the identification and decomposition of this kind of load is still a difficult task. Existing methods for identifying continuously variable loads either require large training data sets or high sampling frequency data for transient feature extraction, which cannot be satisfied by billing smart meters, which also limits the large-scale promotion and application of these methods. application. On the other hand, the charging load of electric bicycles runs for a long time, which may overlap with other electrical loads, which brings great challenges to identification and decomposition. In addition, although there are a few unsupervised and non-invasive electric bicycle charging load detection methods, online detection cannot be achieved.

SUMMARY OF THE INVENTION

Considering the shortcomings of the prior art, in order to further realize the rapid discovery of the charging load of electric bicycles, the present invention combines the non-invasive load monitoring technology, and proposes a non-invasive online rapid detection method of electric bicycle charging load based on local features, It aims to meet the demand for timely and rapid discovery and positioning of electric bicycle charging in actual power consumption scenarios. The invention can accurately and quickly realize on-line detection of electric bicycles, and has broad application prospects in the fields of electric bicycle illegal charging inspection and the like.

In order to solve the above technical problems, the present invention proposes a non-intrusive online fast detection method for electric bicycle charging load, which mainly includes: constructing an electric bicycle charging load template; ; Finally, it is detected whether there is an electric bicycle charging load in the electricity load of the detected user. Specific steps are as follows:

Step 1. Build an electric bicycle charging load template. The steps are as follows:

1-1) Collect the active power and reactive power data of several electric bicycles separately charged with the sampling frequency of 1Hz;

1-2) Data preprocessing: take a time window for the active power and reactive power data collected in step 1-1), and use the state transition removal algorithm to remove the load events in the total active power and total reactive power in the window. Divide; reduce the sampling frequency of active power and reactive power data to 1/30Hz, and use Savitzky-Golay (SG) filtering to reduce the noise in the power signal after the frequency reduction;

1-3) Calculate the difference between the active power and reactive power data preprocessed in step 1-2) to obtain an active power differential signal and a reactive power differential signal;

1-4) Fit the active power differential signal and reactive power differential signal in the constant voltage charging stage as a line segment respectively, and take the maximum point and slope of the fitted line segment as the eigenvector parameters, so as to establish the parameters including active power Electric bicycle charging load template for power differential signal template and reactive power differential signal template;

Step 2. Determine whether there is a suspected electric bicycle charging load in the electricity load of the detected user:

2-1) Collect the active power data and reactive power data of the detected user's home, and calculate the active power differential signal and reactive power at the user's home according to the above steps 1-2) and 1-3). differential signal;

2-2) Match the continuous negative subsequence in the active power differential signal obtained by step 2-1) with the continuous positive subsequence in the reactive power differential signal, and if the matching is successful, perform step 3 , otherwise, repeat step 2;

Step 3: Calculate the distance between the continuous subsequence obtained by the successful matching in Step 2 and the individual charging load template of the electric bicycle constructed in Step 1 as L, if the distance L ≥ the preset distance threshold, return to Step 2, otherwise the user is detected There is an electric bicycle charging load in the electricity load.

Further, the non-invasive electric bicycle charging load online rapid detection method of the present invention, wherein:

For step 1-1), the number of electric bicycles is 10.

For step 1-2), the length of the time window is set to 6 hours, and the step size is set to 10 minutes; the state transition removal algorithm is used to remove the load events in the total active power and total reactive power in the window Removal means that in the power data of the separate charging of the electric bicycle, the opening event of the electric bicycle is removed; in the power data at the user's entrance, the opening, running, and closing events of the electrical load other than the electric bicycle are removed. remove.

For step 2-2), the continuous positive subsequence is the differential signal sequence in which the reactive power differential signal is greater than 0.06, and more than 8 consecutive sampling points are in the same state; the continuous negative subsequence is the active power differential signal A differential signal sequence that is less than -0.1 and has more than 8 consecutive sampling points in the same state; the successful matching means that the consecutive negative active power differential signal subsequences and the consecutive positive reactive power differential signal subsequences overlap in time .

For step 3, the distance L refers to the square sum mean of the difference between the active power differential signal template in the electric bicycle charging load template and the successfully matched active power differential continuous subsequences and the reactive power in the electric bicycle charging load template The average value of the squared sum mean of the differences between the power differential signal template and successive subsequences of successfully matched reactive power differentials.

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

The invention applies the non-intrusive load monitoring technology method to electric bicycle charging load detection, extracts the local features of the electric bicycle charging load, and establishes a non-intrusive electric bicycle charging load online fast detection method based on the local characteristics. The online detection of electric bicycles can be accurately and quickly realized under the condition of the user's interior. This method can meet the demand for timely and rapid discovery and positioning of electric bicycle charging in actual power consumption scenarios, and has broad application prospects in the fields of electric bicycle charging violation inspection and other fields.

Description of drawings

Fig. 1 is the flow chart of the online rapid detection method of the present invention;

Figure 2(a) is a schematic diagram of the active power filtering of the electric bicycle charging stage model;

Figure 2(b) is a schematic diagram of the reactive power filtering of the electric bicycle charging stage model;

Figure 2(c) is a schematic diagram of the active power differential signal after model filtering in the charging stage of the electric bicycle;

Figure 2(d) is a schematic diagram of the reactive power differential signal after model filtering in the charging stage of the electric bicycle;

Figure 3 (a) is a schematic diagram of the detected load event detection result and SG filtering;

Figure 3(b) is a schematic diagram of detected state transition removal and SG filtering;

Figure 4(a) is a graph of 10 electric bicycle charging load slope-active power difference;

Figure 4(b) is a graph of 10 electric bicycle charging load slope-reactive power difference;

Figure 5(a) is a schematic diagram of the active power differential signal and the continuous negative subsequence after SG filtering;

Figure 5(b) is a schematic diagram of the reactive power differential signal and the continuous positive subsequence after SG filtering;

Figure 6(a) is the real value of user power and EBCL power detected online by a certain electric bicycle charging load of user No. 1;

Figure 6(b) is a schematic diagram of the active power differential signal and continuous negative sub-sequence after SG filtering of the online detection of a certain electric bicycle charging load of user No. 1;

Figure 6(c) is a schematic diagram of a reactive power differential signal and a continuous positive subsequence after online detection of a certain electric bicycle charging load of user No. 1 after SG filtering.

Detailed ways

The design idea of the non-intrusive electric bicycle charging load online rapid detection method proposed by the present invention is that, in the research process, it is found that the electric bicycle is different from other electrical loads by collecting the active power and reactive power data of the electric bicycle charging alone. In the constant voltage charging stage, this stage shows the load characteristics of the active power falling gently and the reactive power rising gently. Therefore, the present invention adopts the local characteristics of the constant voltage stage, that is, the difference between the active power and the reactive power amplitude gradually increases. Load feature to detect electric bike charging load.

The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but the following embodiments do not limit the present invention by any means.

The method for realizing the non-intrusive online fast detection method of electric bicycle charging load based on local characteristics of the present invention mainly includes constructing electric bicycle charging load template by using local characteristics of constant voltage stage, and then collecting the active power data and reactive power of the detected user's home. Power data, and based on the calculated active power differential signal and reactive power differential signal at the user's home, determine whether there is suspected electric bicycle charging load in the electricity load of the detected user; finally compare with the electric bicycle charging load template. The pair sum calculation detects whether there is an electric bicycle charging load in the electricity load of the detected user. As shown in Figure 1, the specific steps are as follows:

Step 1. Build an electric bicycle charging load template. The steps are as follows:

1-1) At the sampling frequency of 1Hz, collect active power and reactive power data of 10 electric bicycles separately charged at the same time;

1-2) Data preprocessing: Take a time window for the data collected in step 1-1) for the separate charging of electric bicycles, and use the state transition removal algorithm to remove the load events in the total active power and total reactive power in the window. .

This method adopts an adaptive two-stage event detection method proposed by [Luan W, Liu Z, Liu B, et al. An Adaptive Two-stage Load Event Detection Method for Nonintrusive Load Monitoring.]. The different degrees of electrical load fluctuation in the data adaptively adjust the event detection threshold, and for the step-like events and long transient events with different waveform characteristics, the improved edge detection method and the window-based method combined with moving average and sliding T test are respectively adopted. The detection method can realize the accurate detection and positioning of electrical load events in the total electricity consumption data. The detected state transition removal algorithm is then used to remove it from the total data.

Load events in reactive power are determined from events in active power, and then a state transition removal algorithm is used to remove the detected events. The purpose of state transition removal is to restore the gentle slope trend characteristic of electric bicycle charging load (EBCL) segmented by state transitions of other electrical loads. When a load event is detected in the power signal, it is removed from the total data using equation (1):

where τ is the time index of the analyzed signal, Z is the active and reactive power signals, ΔZ is the value of the active and reactive power of the detected event, and _Lt is the active power signal after state transition removal and Reactive power signal.

Down-sampling and filtering, the sampling frequency of the individual charging data of the electric bicycle is reduced to 1/30Hz, and Savitzky-Golay (SG) filtering is used to reduce the noise in the down-frequency power signal.

In the present invention, in order to preserve the gentle slope characteristics of the electric bicycle and reduce the interference of other electrical loads at the same time, the average value is calculated every 30 sampling points as a new sampling point, that is, the sampling frequency is reduced to 1/30Hz. Since the signal after the state transition is removed still contains the fluctuation in the operation of the electrical load, in order to further reduce the fluctuation and smooth the gentle slope characteristic corresponding to the constant voltage charging stage of the electric bicycle, the present invention adopts the Savitzky-Golay filter (SG filter) to smooth the signal. . SG filtering is a filtering method based on local least squares polynomial fitting with sliding window. This filtering method retains the peak value and width of the original signal while eliminating noise at different frequencies, and is widely used in signals with non-Gaussian noise. Denoise. Compared with filtering methods such as mean filtering and Kalman filtering, SG filtering has better signal shape preservation and denoising performance without loss of resolution.

Given a local symmetric data window n=[l _-m ,l _-m+1 ,...,l ₀ ,...,l _m-1 ,l _m ] of length 2m+1, l _i represents the filtering The active power and reactive power data of a sampling point in the post signal L _t , the signal after SG filtering is:

Where, p<2m, represents the order of the least square polynomial, a _k represents the coefficient of the polynomial, and l' _n is the active power and reactive power signals corresponding to the data window n after SG filtering.

In the present invention, the length of the time window is set to 6 hours, and the step size is set to 10 minutes. The state transition removal algorithm is used to remove the load events in the total active power and total reactive power within the window, and for the power data of the single charging of the electric bicycle, the start event of the electric bicycle is removed.

1-3) Calculate the difference of the power data, and calculate the difference of the pre-processed electric bicycle charging data in step 1-2) to obtain the active power differential signal and reactive power differential signal of the electric bicycle alone charging.

1-4) Use the preprocessed electric bicycle charging power data to construct the electric bicycle charging load template.

和无功功率差分信号模板

图2(a)示出了电动自行车充电阶段模型有功功率滤波示意图；图2(b)示出了电动自行车充电阶段模型无功功率滤波示意图；图2(c)示出了电动自行车充电阶段模型滤波后有功功率差分信号；图2(d)示出了电动自行车充电阶段模型滤波后无功功率差分信号。SG filtering makes the transition from the constant current charging stage to the constant voltage charging stage smoother, and the small fluctuations in the constant voltage charging stage are also smoothed out. The differential signal is calculated for the processed active power data and reactive power data, as shown in Figure 2 ( a), Fig. 2(b), Fig. 2(c) and Fig. 2(d), the power differential signal in the constant voltage charging stage can be fitted as a line segment, with the maximum point of the fitted line segment and The slope is used as the eigenvector parameter of the template. Fit the constant voltage segment of the active power and reactive power data respectively to establish the active power differential signal template of the electric bicycle charging load

and reactive power differential signal mask

Figure 2(a) shows the schematic diagram of active power filtering of the electric bicycle charging stage model; Figure 2(b) shows the schematic diagram of the reactive power filtering of the electric bicycle charging stage model; Figure 2(c) shows the electric bicycle charging stage model Active power differential signal after filtering; Figure 2(d) shows the reactive power differential signal after model filtering in the charging stage of the electric bicycle.

Step 2: Determine whether there is a suspected electric bicycle charging load in the electricity load of the detected user.

2-1) Collect the active power data and reactive power data of the detected user's home, and follow the above steps 1-2) and 1-3), and the data substituted in the process is collected from the user's home. The power data, after downsampling and filtering, for the power data at the user's entrance, the on, running, and off events of electrical loads other than electric bicycles are removed. The active power differential signal and reactive power differential signal at the user's home are obtained by calculating the power data differential.

2-2) Calculate the difference of power data: Match the continuous negative subsequence in the active power differential signal at the user's entrance calculated in step 2-1) with the continuous positive subsequence in the reactive power differential signal.

The present invention sets the active power difference less than -0.1 to be negative, and the reactive power difference greater than 0.06 to be positive, and stipulates that more than 8 consecutive sampling points are in the same state and can be divided into continuous subsequences. In the constant voltage charging phase of the electric bicycle, the active power descending slope and the reactive power ascending slope are synchronized in time. Therefore, it is necessary to match the continuous negative active power differential signal subsequence with the continuous positive reactive power differential signal subsequence, that is, continuous If the negative active power differential signal subsequence and the continuous positive reactive power differential signal subsequence overlap in time, it is considered a match. If the match is successful, go to step 3, otherwise, repeat step 2.

Step 3: Match with the electric bicycle charging load template, by calculating the distance between the matched continuous subsequences and the electric bicycle charging load template, if the distance is less than the preset distance threshold, it is considered to match the template.

For the time-synchronized (ie successfully matched) active power differential signal continuous negative subsequence and reactive power differential signal continuous positive subsequence, calculate the distance from the electric bicycle charging load template, where the distance refers to the electric bicycle The square sum mean of the difference between the active power differential signal template in the charging load template and the successive subsequences of successfully matched active power differentials and the reactive power differential signal template in the electric bicycle charging load template and the successfully matched reactive power differential The average value of the square and mean of the differences between consecutive subsequences, as shown in formula (3):

Among them, x is the electric bicycle charging load template signal, y is the successfully matched power differential continuous subsequence, x1 is the maximum value point of the template, _y1 is the maximum value point _of the continuous subsequence, and n is the differential signal continuous subsequence. The number of data points from the maximum point to the termination point of the subsequence. It should be noted that in order to avoid the situation where the maximum point in the continuous subsequence is in the last digit and is close to the highest point of the template, n>10, ΔS is required. Active signal subsequence distance and active signal subsequence distance. Take the average of the distance of the active signal subsequence and the distance of the reactive signal subsequence as the distance between the matching subsequence and the template. If it is less than the distance threshold set by the heuristic, it is considered that the subsequence matches the template, and there is a constant voltage charging stage of the electric bicycle. , that is, it is detected that there is an electric bicycle charging load in the user's electricity load; if it is greater than or equal to the threshold, it is considered that the subsequence does not match the template, and there is no electric bicycle constant voltage charging stage.

Research Material Example 1:

Figure 3(a) and Figure 3(b) respectively show the results of load event detection and state transition removal for the data at the user entrance of a certain period of time, and also show the comparison images before and after SG filtering, it can be seen that The load event detection algorithm can accurately detect events in the total electricity consumption data.

最低点：-1.2836；

斜率：0.0100；

最高点：0.5548；

斜率：-0.0054。The electric bicycle charging load template is constructed by using the active power data and reactive power data of 10 electric bicycles individually charged. The 10 electric bicycle charging load model parameters shown in the table above are obtained. According to the data in Table 1, the model diagram of 10 electric bicycle charging loads is drawn. Figure 4(a) shows the slope-active power difference diagram of 10 electric bicycle charging loads, and Figure 4(b) shows 10 electric bicycle charging loads. Slope-Reactive Power Differential Plot. It can be seen that the model of No. 5 lead-acid battery obviously deviates from the other 9 models, so it is eliminated, and the average value of the remaining 9 models is obtained to obtain the EBCL template, namely

Lowest: -1.2836;

slope: 0.0100;

Highest point: 0.5548;

Slope: -0.0054.

Table 1 10 electric bicycle charging load model parameters

Calculate the differential signal for the active power and reactive power data at the entrance, and match the continuous negative subsequence in the active power differential signal with the continuous positive subsequence in the reactive power differential signal, as shown in Figure 5(a) and As shown in 5(b), it can be seen that the deepened two subsequences around 00:00 match. Finally, the distance threshold is set to 0.25. Figure 6(a), Figure 6(b) and Figure 6(c) show the online detection results of a certain electric bicycle charging load for user No. 1, where Figure 6(a) is 1 Figure 6(b) is a schematic diagram of the active power differential signal and continuous negative sub-sequence after online detection of a certain electric bicycle charging load of user No. 1 after SG filtering; Figure 6(c) is a schematic diagram of a reactive power differential signal and a continuous positive subsequence after online detection of a certain electric bicycle charging load of user No. 1 after SG filtering.

Research Material Example 2:

At the same time, the active power and reactive power data at the entrance of 7 users for a total of 411 days were tested online for the charging load of electric bicycles. Table 2 shows the experimental results of the 7 users. It can be seen that there is almost no leakage of charging load. At the same time, it basically realizes the online rapid detection from 20 minutes to 45 minutes after the constant voltage charging starts.

Table 2 Results of online rapid detection of electric bicycle charging load

According to the above implementation, the present invention can quickly detect the electric bicycle charging load with high accuracy without intruding into the user's interior according to the summed up electric bicycle charging load template. This method has great practical feasibility in the application of high-efficiency detection of illegal charging of electric bicycles, and can meet the needs of rapid discovery and positioning of illegal charging behaviors of electric bicycles in actual power consumption scenarios. The field has broad application prospects.

Although the present invention has been described above in conjunction with the accompanying drawings, the present invention is not limited to the above-mentioned specific embodiments, which are merely illustrative rather than restrictive. Under the inspiration of the present invention, many modifications can be made without departing from the spirit of the present invention, which all belong to the protection of the present invention.

Claims (5)

1. A non-intrusive type electric bicycle charging load online rapid detection method is characterized by comprising the following steps:

step 1, constructing a charging load template of the electric bicycle, comprising the following steps:

1-1) acquiring active power and reactive power data of independent charging of a plurality of electric bicycles with the sampling frequency of 1 Hz;

1-2) data preprocessing: taking a time window for the active power and reactive power data acquired in the step 1-1), and removing load events in the total active power and the total reactive power in the window by adopting a state conversion removal algorithm; reducing the sampling frequency of the active power data and the reactive power data to 1/30Hz, and adopting Savitzky-Golay (SG) filtering to reduce the noise in the power signals after frequency reduction;

1-3) calculating difference of the active power and reactive power data preprocessed in the step 1-2) to obtain an active power difference signal and a reactive power difference signal;

1-4) respectively fitting the active power differential signal and the reactive power differential signal in a constant voltage charging stage into a line segment, and taking the maximum point and the slope of the fitted line segment as characteristic vector parameters, thereby establishing an electric bicycle charging load template comprising an active power differential signal template and a reactive power differential signal template;

Step 2, judging whether the charging load of the electric bicycle exists in the power load of the detected user:

2-1) acquiring active power data and reactive power data of the user to be detected at the user position, and calculating to obtain an active power differential signal and a reactive power differential signal at the user position according to the step 1-2) and the step 1-3);

2-2) matching continuous negative subsequences in the active power differential signals and continuous positive subsequences in the reactive power differential signals, which are obtained by calculation in the step 2-1), of the user's house, if the subsequences are successfully matched, executing the step 3, and otherwise, repeatedly executing the step 2;

and 3, step 3: and (3) calculating the distance between the continuous subsequence successfully matched in the step (2) and the single charging load template of the electric bicycle constructed in the step (1) to be L, if the distance L is more than or equal to a preset distance threshold value, returning to the step (2), otherwise, detecting that the charging load of the electric bicycle exists in the electric load of the user.

2. The method for the on-line rapid detection of the charging load of the non-invasive electric bicycle according to claim 1, wherein for step 1-1), the number of the electric bicycles is 10.

3. The method for the on-line fast detection of the charging load of the non-invasive electric bicycle according to claim 1, wherein for the steps 1-2), the length of the time window is set to 6 hours, and the step length is set to 10 minutes;

The method comprises the following steps that a load event in total active power and total reactive power in a window is removed by adopting a state conversion removal algorithm, namely, a starting event of the electric bicycle is removed in power data of independent charging of the electric bicycle; in the power data at the user entry, the on, running, off events of the electrical loads other than the electric bicycle are removed.

4. The method for rapidly detecting the charging load of the non-invasive electric bicycle on line according to claim 1, wherein for step 2-2), the continuous positive subsequence is a differential signal sequence in which the reactive power differential signal is greater than 0.06 and more than 8 continuous sampling points are in the same state; the continuous negative subsequence is a differential signal sequence with active power differential signals smaller than-0.1 and more than 8 continuous sampling points in the same state; the successful matching means that the continuous negative active power differential signal subsequence and the continuous positive reactive power differential signal subsequence overlap in time.

5. The method for non-invasive online rapid detection of charging load of electric bicycle according to claim 1, wherein for step 3, the distance L is an average of a mean of a sum of squares of differences between the active power differential signal template in the charging load template of electric bicycle and the successfully matched active power differential continuous subsequence, and a mean of a sum of squares of differences between the reactive power differential signal template in the charging load template of electric bicycle and the successfully matched reactive power differential continuous subsequence.

Images