Intelligent electric energy meter data acquisition method and system based on Internet of Things

Abstract

The invention relates to the technical field of data processing, in particular to an intelligent electric energy meter data collection method and system based on the Internet of Things, and the method comprises the steps: collecting the operation parameters of an electric energy meter, and enabling the operation parameters to comprise current, voltage and load power; and performing smoothing processing on the operation parameters by using a weighted moving average method, inputting the smoothed operation parameters into a prediction model to obtain predicted electricity consumption, and performing abnormal early warning according to a comparison result of the predicted electricity consumption output by the prediction model and actual electricity consumption. According to the method, the operation parameters are smoothed by using the weighted moving average method and the credibility, so that the data quality is improved; and the abnormal early warning is realized by calculating the abnormal coefficient of the predicted electricity consumption and the actual electricity consumption, so that the reliability and the prediction precision of the electric energy meter data are effectively improved, the equipment abnormality can be found in advance, the maintenance cost is reduced, and the stable operation of the power system is ensured.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology. More specifically, the present invention relates to a method and system for collecting data from a smart energy meter based on the Internet of Things. Background Art

[0002] With the rapid development of IoT technology, smart energy meters, as a key component of smart grids, enable precise energy measurement, remote communication, data analysis, and user energy consumption management. The introduction of IoT technology has enabled smart energy meters to acquire real-time data, perform remote control, monitor status, and perform big data analysis, enabling remote collection, storage, and analysis of user electricity usage data. This not only improves the reliability and real-time nature of data transmission but also, through the integration of big data and artificial intelligence technologies, provides value-added services such as intelligent load management, power quality analysis, and abnormal power usage warnings.

[0003] In related technologies, for example, Chinese patent application document with publication number CN119179900A discloses a method and system for collecting parameters of smart electricity meters based on the Internet of Things. The method includes: obtaining a predicted electricity increment at the current moment based on a prediction model; comparing a real-time error with an error threshold to obtain an operating status, where the real-time error is the absolute value of the difference between the predicted electricity increment and the measured electricity increment; and storing the electricity meter reading if the operating status is normal; otherwise, issuing an early warning. The technical solution of this application can accurately determine the operating status of the electricity meter, ensuring that the electricity meter can collect accurate electricity consumption.

[0004] The above method can determine the operating status of the target electricity meter by comparing the real-time error and the error threshold. However, since the electricity meter is usually installed in the power system, there may be high-voltage lines, transformers, motors and other equipment around it, which will generate electromagnetic interference, or wireless communication equipment may also introduce high-frequency interference, thereby causing noise to the collection of the operating parameters of the electricity meter. When the prediction model uses the operating parameters with noise for prediction, it will lead to the problem of low prediction accuracy. Summary of the Invention

[0005] The present invention provides a data collection method and system for smart electric energy meters based on the Internet of Things, aiming to solve the problem in related technologies that noise may affect the collection of operating parameters of electric energy meters, resulting in low accuracy of prediction results when the prediction model uses operating parameters with noise for prediction.

[0006] In a first aspect, the present invention provides a method for collecting data from an intelligent electric energy meter based on the Internet of Things, comprising: collecting operating parameters of the electric energy meter, wherein the operating parameters include current, voltage, and load power; smoothing the operating parameters using a weighted moving average method, inputting the smoothed operating parameters into a prediction model to obtain predicted power consumption, and issuing an abnormality warning based on a comparison result between the predicted power consumption output by the prediction model and the actual power consumption; wherein smoothing the operating parameters includes: obtaining a target sequence of the electric energy meter's operating parameters at consecutive moments, calculating the noise level at any moment, wherein the noise level is positively correlated with the volatility of the operating parameters; calculating the credibility at any moment, wherein the credibility is related to the degree of difference between the target vector at that moment and the standard vector of each operating parameter at that moment, and also to the noise level at that moment, wherein the target vector reflects the changing trend of each operating parameter at that moment; and weighting the time series weighting factor in the weighted moving average method using the credibility at each moment to obtain the smoothed operating parameters at each moment. By comparing the difference between the predicted power consumption and the actual power consumption, combined with the smoothed operating parameters, abnormal power consumption fluctuations can be more accurately identified. When anomalies occur in the operating parameters of the electricity meter, the smoothed data is more accurate, which can effectively reduce false alarms caused by noise and improve the accuracy of abnormal warnings.

[0007] Furthermore, an abnormal warning is performed, including: calculating the absolute value of the difference between the predicted power consumption and the actual power consumption as an abnormal coefficient; if the abnormal coefficient is greater than the abnormal threshold, an alarm is issued to prompt the staff to perform maintenance.

[0008] Furthermore, the degree of noise at any moment is calculated, including: calculating the degree of difference between the operating parameters at any moment in the target sequence and the average value of the operating parameters at all moments, and the variance of the operating parameters at all moments in the target sequence; the volatility of the operating parameters is positively correlated with the degree of difference and the variance. By comparing the operating parameters at each moment in the target sequence with the average value of all moments, it is possible to accurately reveal whether the operating parameters at that moment deviate from the overall trend. If the degree of difference at a certain moment is large, it means that the operating parameters at that moment may be affected by strong noise, further enhancing the sensitivity of noise detection. In addition, combined with variance calculation, the fluctuation range of the data can be quantified, thereby judging the stability of the data and the degree of influence of noise.

[0009] Furthermore, the target vector is ;in, Indicates the Moment The changing trend of various operating parameters, Indicates the number of different operating parameters.

[0010] Furthermore, obtaining a standard vector for each operating parameter at that moment includes: determining the power consumption period at that moment as the target period, where power consumption periods include peak periods, normal periods, and off-peak periods; obtaining historical target vectors for each moment in the target period from historical data, selecting all historical target vectors that have the same changing trend as any operating parameter in the target vector at that moment, and averaging all of these historical target vectors as the standard vector for each operating parameter at that moment. By selecting appropriate historical data based on the power consumption period and calculating the average of historical target vectors with similar changing trends to the target vector at the current moment as the standard vector, the model's adaptability to different time periods, data stability, and prediction accuracy can be significantly improved.

[0011] Furthermore, the changing trend of each operating parameter at that moment is obtained to satisfy the following relationship: Where, Indicates the The changing trend of operating parameters at all times, Indicates the The parameter value of the running parameter at all times, Indicates the The parameter value of the running parameter at all times, Representing a symbolic function; by using symbolic functions to determine trend changes, calculations become significantly simpler, reducing computational complexity while enabling rapid response to real-time data. This simplified calculation process is particularly useful for large-scale power data monitoring systems, which can process large amounts of data and determine trends in a short period of time.

[0012] Further, Credibility of the moment The calculation formula is: Where, Indicates the The target vector at time t, Indicates the Moment A standard vector of operating parameters, For the The noise level at the time, Indicates the number of operating parameters, Expressed as a natural constant By calculating the difference between the target vector and the standard vector and combining the degree of noise, the system can dynamically evaluate the The credibility of the data at that moment. When the target vector differs significantly from the standard vector, the credibility is low; conversely, when the difference is small, the credibility is high. This method can reflect the reliability of the data in a timely manner, especially in the face of sudden fluctuations or anomalies, and can provide an immediate assessment of data quality.

[0013] Furthermore, the smoothed operating parameters at each moment are calculated using the following formula: Where, Indicates the Moment The final parameter value after smoothing the running parameter value, represents the size of the sliding window, Indicates the first Timing weighting factor of the moment operation parameter, Indicates the first Credibility of the moment, Indicates the first The operating parameter value at the moment.

[0014] Furthermore, the prediction model is an LSTM model or a GUR model.

[0015] In a second aspect, the present invention further provides an Internet of Things-based smart energy meter data acquisition system, comprising a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to implement any of the above-mentioned Internet of Things-based smart energy meter data acquisition methods.

[0016] Beneficial effects: By combining IoT technology, the operating parameters of the electricity meter are collected and smoothed using the weighted moving average method and credibility, the fluctuation or noise in the operating parameters is reduced and the data quality is improved. The LSTM or GRU model is used to predict electricity consumption, and abnormal warnings are achieved by calculating the abnormal coefficient between the predicted and actual electricity consumption. The overall solution effectively improves the reliability and prediction accuracy of electricity meter data, can detect equipment abnormalities in advance, reduce maintenance costs, and ensure the stable operation of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 FIG. 4 is a flow chart schematically illustrating smoothing of operating parameters according to an embodiment of the present invention. DETAILED DESCRIPTION

[0018] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0019] like Figure 1 As shown, S101: collecting operating parameters of the electric energy meter.

[0020] In one embodiment, the operating parameters of the electric energy meter at each moment are obtained, where the operating parameters include current, voltage, and load power, and the operating parameters are parameters directly collected by the electric energy meter. In other embodiments, the operating parameters of the electric energy meter also include power consumption at each moment, where the power consumption at each moment is the difference between the power consumption at that moment and the power consumption at the previous moment.

[0021] S102: Calculate the noise level at each moment.

[0022] In one embodiment, an electric energy meter is typically installed in an electric power system, and there may be high-voltage lines, transformers, motors, or wireless communication equipment around it. These devices generate various forms of electromagnetic interference during operation, including radio frequency interference, electromagnetic radiation interference, and power frequency interference. These interference signals may be superimposed on the measurement signal of the electric energy meter, causing deviations or noise in the collected parameters, thereby affecting the measurement accuracy. When the parameters collected by the electric energy meter are affected by electromagnetic interference and errors occur, the prediction model trained based on these parameters may be affected, resulting in a decrease in the reliability of the prediction results. Noise data often deviates significantly from the main distribution trend of the data. Compared with normal data points, they may show extreme values or different patterns. Therefore, it is necessary to calculate the noise level of the collected parameters at each moment based on this characteristic. In summary, the noise level at each moment is positively correlated with the volatility of the operating parameters.

[0023] In one embodiment, calculating the noise level at each moment includes: obtaining a target sequence of the electricity meter's operating parameters at consecutive moments. Taking the current moment t as an example, obtaining a time series consisting of multiple moments before the current moment t. Since the operating parameters collected at each moment include voltage, current, and load power, multiple corresponding operating parameter sequences can be obtained, including a voltage sequence, a current sequence, and a load power sequence. At this point, the noise level at any moment can be calculated using the following formula: Where, Indicates the The noise level at the time, Indicates the number of sequences, Indicates the The first parameter values, Indicates the The average value of all parameter values in the series, Indicates the The variance of all parameter values within a series.

[0024] in, Indicates the The first The parameter value and The difference between the mean values of all parameter values in a sequence, The larger the value, the The greater the fluctuation between a parameter value and the whole, the higher the noise level of the parameter value. The larger the value, the The greater the dispersion of all parameter values in a sequence, the higher the noise level of the parameter value.

[0025] S103: Calculate the credibility at any moment.

[0026] In one embodiment, when the user uses electricity, the changing trends of current, voltage, and load power should be the same. For example: when the current at a certain moment increases compared to the previous moment, then the load power at that moment should also increase, and the power consumption at that moment should also increase; if the current increases at this time, and the load power or power consumption decreases or remains unchanged, it means that the operating parameters collected at that moment may be abnormal, and the credibility of the operating parameters collected at that moment is low. Specifically, the credibility of any moment can be calculated based on the degree of difference between the target vector of the operating parameters at the current moment and the standard vector of each operating parameter at that moment, and the noise level at that moment. The greater the degree of difference and the noise level at that moment, the lower the credibility of the operating parameters at that moment. Among them, the target vector reflects the changing trend of each operating parameter at that moment. Among them, to obtain the changing trend of each operating parameter at that moment, the following relationship must be satisfied: Where, Indicates the The changing trend of operating parameters at all times, Indicates the The parameter value of the running parameter at all times, Indicates the The parameter value of the running parameter at all times, Represents a symbolic function.

[0027] It should be noted that the target vector of the current operating parameters is obtained, and the target vector is ;in, Indicates the Moment The changing trend of various operating parameters, The number of different operating parameters is represented by , and then the standard vectors for each operating parameter at that moment are selected based on the current power consumption period, where the power consumption period includes peak period, normal period, and off-peak period. The reason for this is that due to the differences in power consumption period, when using target vectors of operating parameters from different power consumption periods for comparison, the comparison results are inaccurate, which can easily affect the calculation results. For example, when the target vector of operating parameters is collected during the peak period, it is compared with the standard vector of operating parameters collected during the off-peak period. Since the voltage, current, and load power during the off-peak period are much lower than those during the peak period, there will be a large difference between the parameters, resulting in errors in the final calculated result. Therefore, it is necessary to determine the power consumption period at the current moment and use this to obtain the standard vector of the operating parameters at that moment.

[0028] For example, peak hours are 6:00 AM to 9:00 AM and 6:00 PM to 10:00 PM, as these are the peak hours due to high electricity demand, such as from water heaters, kitchen appliances (rice cookers, induction cookers, microwave ovens), lighting, and air conditioning. Off-peak hours are 2:00 PM to 6:00 PM and 10:00 PM to 6:00 AM, when no one is home or appliances are used less frequently, resulting in lower electricity loads. The remaining hours are considered normal. It should be noted that selecting the time period corresponding to the collection time allows for comparison of changes corresponding to the collection time within the same time period, providing a more meaningful reference and improving the accuracy of subsequent reliability calculations.

[0029] As mentioned above, the division of electricity usage into peak, normal, and off-peak hours helps identify typical patterns of electricity usage within different time periods. By selecting historical data similar to the current time period, the system can more accurately match patterns within specific time periods. This allows the model to better adapt to temporal variations, thereby improving its ability to predict electricity usage and detect anomalies across different time periods.

[0030] In one embodiment, obtaining a standard vector for each operating parameter at any moment includes: determining the power consumption period at the moment as a target period, where the power consumption period includes a peak period, a normal period, and a valley period; obtaining a historical target vector for each moment in the target period in historical data, and selecting all historical target vectors with the same change trend as any operating parameter in the target vector at the moment, and the average value of all the historical target vectors as the standard vector for each operating parameter at the moment.

[0031] In another embodiment, since there may be abnormal historical target vectors among all historical target vectors, it is necessary to cluster all historical target vectors that have the same change trend as any operating parameter in the target vector at that moment to obtain two categories, and use the average value of the historical target vectors in the largest category as the standard vector for each operating parameter at that moment. The reason is that since there are often fewer abnormal data during the operation of the electric energy meter, clustering can effectively distinguish abnormal data from normal data. Usually, normal data will form a major cluster, while abnormal data will belong to a small cluster or outlier. Selecting the historical target vector with the largest category as the standard vector can ignore the impact of a small number of abnormal data and prevent them from interfering with the final result.

[0032] In one embodiment, there is provided Credibility of the moment The calculation formula is: Where, Indicates the The target vector at time t, Indicates the Moment A standard vector of operating parameters, For the The noise level at the time, Indicates the number of operating parameters, Expressed as a natural constant An exponential function with base . Indicates the The target vector at time The difference between the standard vectors of the operating parameters is greater. The more inconsistent the target vector and the standard vector are, the The lower the credibility of the moment; the smaller the difference, the The higher the consistency of the target vector and the standard vector at the moment, the The higher the credibility of the moment. Noise level at the time By weighted processing, the greater the noise level, the lower the credibility, which improves the accuracy of the calculated credibility.

[0033] S104: Smoothing the operating parameters.

[0034] In one embodiment, the operating parameters can be smoothed using a weighted moving average method. The basic concept of the weighted moving average method is that recent observations are more important than older observations, so more recent data is given a greater weight and older data is given a smaller weight. By weighting and averaging the data, the weighted moving average can reduce the impact of sudden changes and smooth the data.

[0035] In one embodiment, the reliability of the operating parameters at each moment is used to weight the time series weighting factor in the weighted average moving average method to calculate the smoothed operating parameters at each moment. The calculation formula is: Where, Indicates the Moment The final parameter value after smoothing the running parameter value, Indicates the size of the sliding window. The size of the sliding window is 3. Indicates the first The timing weighting factor of the moment-to-moment operating parameter, which reflects the temporal importance of the data (e.g., newer data may be assigned a higher timing weighting factor), Indicates the first Credibility of the moment, Indicates the first The operating parameter value at the moment. The time represents the current moment, that is, time, is the latest collected operating parameter in the sliding window, then the calculation formula of the timing weighting factor of the operating parameter is: = At this point, the above method can be used to calculate the smoothed operating parameters at each moment.

[0036] According to the above steps, by introducing the credibility of each moment, high-credibility data contributes more to the results, while low-credibility data contributes less, avoiding the dominant role of erroneous data in the smoothed results. This can reduce the impact of inaccurate data on the final calculation results, reduce the interference of outliers, and make the smoothed results more representative. Furthermore, by removing noise through the weighted moving average method, the random fluctuations of the electricity meter operating parameters can be effectively reduced. This makes the data input into the prediction model more stable and accurate, which can significantly improve the accuracy of predicted electricity consumption.

[0037] S105: Determine whether there is any abnormality in the electric energy meter.

[0038] In one embodiment, the operating parameters are smoothed using a weighted moving average method, and the smoothed operating parameters are input into a prediction model to obtain predicted power consumption. An abnormality warning is performed based on the comparison result between the predicted power consumption output by the prediction model and the actual power consumption, wherein the prediction model is an LSTM model or a GUR model. Specifically, the absolute value of the difference between the predicted power consumption and the actual power consumption is calculated as the abnormality coefficient; if the abnormality coefficient is greater than the abnormality threshold, it indicates that there is a large error between the actual power consumption collected by the electricity meter and the predicted power consumption, which also indicates that there is an abnormality in the electricity meter, and an alarm is issued to prompt the staff to repair it. The empirical value of the abnormality threshold is 1. In other embodiments, the empirical value of the abnormality threshold can be 1.5 or 0.6, etc., which can be adjusted according to the specific implementation situation.

[0039] In another embodiment, a graded warning can be issued based on the magnitude of the abnormality coefficient. Specifically, the warning levels can be divided into normal, low risk, and high risk, and the corresponding alarm is issued according to the warning level of the abnormality coefficient. For example, the normal warning range is less than 0.6, the low risk warning range is greater than or equal to 0.6 and less than 1.5, and the high risk warning range is greater than or equal to 1.5 and less than or equal to 2.

[0040] The present invention also provides an IoT-based smart energy meter data collection system. The system includes a processor and a memory, wherein the memory stores computer program instructions. When the computer program instructions are executed by the processor, the method for collecting smart energy meter data based on the IoT according to the first aspect of the present invention is implemented.

[0041] The system further includes other components well known to those skilled in the art, such as a communication bus and a communication interface. The configuration and functions of these components are known in the art and thus will not be described in detail here.

[0042] In the present invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, the computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store the required information and can be accessed by an application, module, or both. Any such computer storage medium can be part of, accessible to, or connectable to a device. Any application or module described in the present invention can be implemented using computer-readable / executable instructions stored or otherwise retained on such a computer-readable medium.

[0043] The above-described embodiments merely illustrate several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, and all such variations and improvements fall within the scope of protection of the present invention.

Claims

1. A method for collecting data from smart electric energy meters based on the Internet of Things, characterized in that: include: Collecting the operating parameters of the electric energy meter, including current, voltage and load power; The operating parameters are smoothed using a weighted moving average method, the smoothed operating parameters are input into a prediction model to obtain predicted power consumption, and an abnormality warning is issued based on a comparison result between the predicted power consumption output by the prediction model and the actual power consumption; The step of smoothing the operating parameters includes: Obtaining a target sequence of operating parameters of the electric energy meter at consecutive moments and calculating the noise level at any moment, wherein the noise level is positively correlated with the volatility of the operating parameters; Calculate the credibility at any moment, which is related to the degree of difference between the target vector at that moment and the standard vector of each operating parameter at that moment, and also to the noise level at that moment, where the target vector reflects the changing trend of each operating parameter at that moment; The credibility of each moment is used to perform weighted processing on the time series weighting factor in the weighted moving average method to obtain the smoothed operating parameters at each moment.

2. The method for collecting data of smart electric energy meters based on the Internet of Things according to claim 1, characterized in that: Provide abnormal warnings, including: Calculating the absolute value of the difference between the predicted power consumption and the actual power consumption as an abnormality coefficient; If the abnormal coefficient is greater than the abnormal threshold, an alarm is issued to prompt the staff to perform maintenance.

3. The method for collecting data of smart electric energy meters based on the Internet of Things according to claim 1, characterized in that: Calculate the noise level at any moment, including: Calculate the difference between the operating parameters at any moment in the target sequence and the average value of the operating parameters at all moments, and the variance of the operating parameters at all moments in the target sequence; The volatility of the operating parameters is positively correlated with both the degree of difference and the variance.

4. The method for collecting data of smart electric energy meters based on the Internet of Things according to claim 1, characterized in that: The target vector is ;in, Indicates the Moment The changing trend of various operating parameters, Indicates the number of different operating parameters.

5. The method for collecting data of smart electric energy meters based on the Internet of Things according to claim 1, characterized in that: Get the standard vector of each operating parameter at that moment, including: Determine the electricity consumption period at the moment as the target period, where the electricity consumption period includes peak period, normal period and off-peak period; Obtain the historical target vectors at each moment in the target period in the historical data, and select all historical target vectors with the same change trend as any operating parameter in the target vector at that moment, and use the average value of all the historical target vectors as the standard vector of each operating parameter at that moment.

6. The method for collecting data of smart electric energy meters based on the Internet of Things according to claim 1, characterized in that: Obtain the change trend of each operating parameter at that moment, satisfying the following relationship: ； Where, Indicates the The changing trend of operating parameters at all times, Indicates the The parameter value of the running parameter at all times, Indicates the The parameter value of the running parameter at all times, Represents a symbolic function.

7. The method for collecting data of smart electric energy meters based on the Internet of Things according to claim 1, characterized in that: No. Credibility of the moment The calculation formula is: ； Where, Indicates the The target vector at time t, Indicates the Moment A standard vector of operating parameters, For the The noise level at the time, Indicates the number of operating parameters, Expressed as a natural constant An exponential function with base .

8. The method for collecting data of smart electric energy meters based on the Internet of Things according to claim 1, characterized in that: Calculate the smoothed operating parameters at each moment using the following formula: ； Where, Indicates the Moment The final parameter value after smoothing the running parameter value, represents the size of the sliding window, Indicates the first Timing weighting factor of the moment operation parameter, Indicates the first Credibility of the moment, Indicates the first The operating parameter value at the moment.

9. The method for collecting data of smart electric energy meters based on the Internet of Things according to claim 1, characterized in that: The prediction model is an LSTM model or a GUR model.

10. A smart energy meter data acquisition system based on the Internet of Things, comprising a processor and a memory, characterized in that: The memory stores a computer program, and the processor executes the computer program to implement the smart electricity meter data collection method based on the Internet of Things as described in any one of claims 1 to 9.

Images