A method and system for monitoring electricity consumption of smart electric meter
Through deep learning algorithms and signal processing technology, real-time monitoring and prediction of power consumption equipment by smart meters is realized, solving the problem that existing smart meters cannot predict and deal with abnormal power consumption, and improving power consumption safety and efficiency.
Patent Information
- Application Number
- CN202410133609.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-01-31
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2044-01-31
AI Technical Summary
Existing smart meters cannot monitor and predict the power consumption of factory equipment in real time, and lack the ability to judge and handle abnormal power consumption, resulting in low power consumption safety and efficiency.
Deep learning algorithm is used to preprocess the historical and current data in the smart meter database, predict the power of power consumption equipment in the next hour, and monitor the load power in real time through signal conversion and load power calculation modules, judge abnormal power consumption, and output Corresponding alarm signals or control commands.
It improves the intelligence level of smart meters, realizes dynamic adjustment and optimization of power consumption equipment, enhances power safety, promptly detects and deals with abnormal power consumption, avoids wasting or exceeding the standard power consumption, and reduces power consumption costs.
Smart Images

Figure CN117990976B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of electric power monitoring, and in particular to a method and system for monitoring electric power consumption of a smart electric meter. Background Art
[0002] Plant equipment is an important part of industrial production, and its power safety and efficiency directly affect the quality and benefits of industrial production. With the development of industrial production scale, automation and intelligence, the types and quantity of plant equipment are increasing, and its power demand and power characteristics are becoming more complex and changeable. In order to effectively monitor and manage the power consumption of plant equipment, it is necessary to establish a complete smart meter power monitoring system to realize the functions of power prediction, load power calculation, abnormal power detection, power efficiency analysis, power suggestions or optimization plans for plant equipment.
[0003] At present, my country has installed smart meters in some factories, but the functions of these smart meters are not perfect enough, and there are some problems and shortcomings, which are mainly manifested in the following aspects:
[0004] (1) Smart meters have weak data collection and analysis capabilities and are unable to predict the power consumption of plant equipment. Currently, most smart meters can only collect and store historical and current data of plant equipment, and lack the function of predicting the power consumption of electrical equipment within the next hour. Since it is impossible to predict future power demand and make reasonable adjustments or controls based on the predicted values, it may lead to power overload or underuse, affecting the normal operation of plant equipment.
[0005] (2) Smart meters are not capable of detecting and handling abnormalities, and are unable to detect and handle abnormal power usage in a timely manner. Currently, most smart meters can only compare the load power with the preset power and determine whether there is abnormal power usage based on the comparison results. They lack the ability to determine the type of abnormal power usage and output corresponding alarm signals or control instructions. Since it is impossible to identify the type of abnormal power usage, such as whether other devices are connected, whether a fault has occurred, and whether the power usage data has been maliciously tampered with, it is impossible to take corresponding measures, which may lead to safety hazards or economic losses.
[0006] To sum up, there are still many problems and shortcomings in the current smart meter electricity consumption monitoring system of factory equipment in my country, and it is urgently needed to be improved and innovated. Summary of the invention
[0007] In view of the above problems, the present invention provides a smart meter electricity consumption monitoring method for realizing real-time monitoring and analysis of the smart meter electricity consumption, thereby improving electricity safety and efficiency. The method comprises the following steps:
[0008] Step 1: The data prediction module collects historical data and current data in the smart meter database, uses a deep learning algorithm to predict the power of the electrical equipment in the next hour, and uses the predicted value as the preset power;
[0009] Step 2: The signal conversion module collects the voltage signal and the current signal of the load circuit, and converts the voltage signal and the current signal into a digital signal;
[0010] Step 3, the load power calculation module calculates the load power according to the converted digital signal;
[0011] Step 4: The abnormality detection module compares the load power with the preset power, and determines whether there is abnormal power usage based on the comparison result, and outputs a corresponding alarm signal or control instruction based on the abnormal power usage.
[0012] This invention solves the technical problem in the prior art that smart meters are unable to monitor and predict the power consumption of electrical equipment in real time. It adopts a smart meter power consumption monitoring method based on a deep learning algorithm, which can predict the power of electrical equipment in the next hour, and judge whether there is abnormal power consumption based on the difference between the actual load power and the preset power, and output corresponding alarm signals or control instructions.
[0013] The invention has the following beneficial effects: it improves the intelligence level of smart meters and can dynamically adjust and optimize the power consumption of electrical equipment; it improves the safety of electricity use and can promptly detect and handle abnormal power consumption to prevent damage to electrical equipment or accidents such as fire; it improves electricity efficiency and saves electricity costs and can reasonably arrange and adjust electricity plans based on prediction results to avoid waste or excessive electricity consumption.
[0014] In the best implementation, the technical problem that needs to be solved is: how to effectively preprocess the collected data to improve the training effect and prediction accuracy of the deep learning model.
[0015] In the best implementation, the technical solution adopted is:
[0016] Step 11, collecting historical data and current data in the smart meter database;
[0017] Step 12, preprocessing the collected data, including removing noise, outliers and missing values;
[0018] Step 13, using the preprocessed data to train a deep learning model and evaluate the performance of the model;
[0019] Step 14: Use the trained deep learning model to predict the power of the electrical equipment in the next hour, and use the predicted value as the preset power.
[0020] Under optimal implementation conditions, the beneficial effects produced are: it can effectively improve the quality and availability of data, reduce data noise and anomalies, increase data integrity and consistency, thereby improving the training effect and prediction accuracy of the deep learning model, and improving the reliability and accuracy of the smart meter electricity consumption monitoring method.
[0021] In a preferred implementation, the technical problem to be solved is: how to effectively filter the digital signal to improve the calculation accuracy and stability of the load power.
[0022] In the best implementation, the technical solution adopted is:
[0023] Step 31, receiving digital signals and storing them in two arrays, respectively representing digital voltage signals and digital current signals;
[0024] Step 32, filtering the digital signals in the two arrays to remove noise and interference and obtain smoothed digital signals;
[0025] Step 33, calculate the load power according to the smoothed digital signal, and the calculation formula is as follows:
[0026]
[0027] Where P is the load power, T is the cycle time, v(t) is the smoothed digital voltage signal, i(t) is the smoothed digital current signal, and t is the variable value of the time interval length in calculating the instantaneous power.
[0028] Under the best implementation conditions, the beneficial effects produced are: it can effectively improve the quality and stability of digital signals, reduce the noise and interference of digital signals, increase the smoothness and continuity of digital signals, and thus improve the calculation accuracy and stability of load power.
[0029] In the preferred implementation, the technical problem to be solved is: how to effectively judge and handle abnormal power consumption and give corresponding alarm signals or control instructions.
[0030] In the best implementation, the technical solution adopted is:
[0031] Step 41, setting the judgment criteria and thresholds for abnormal power usage;
[0032] Step 42, comparing the load power with the preset power to obtain a deviation value and a deviation ratio between the load power and the preset power;
[0033] Step 43, judging whether there is abnormal power consumption according to the judgment standard and threshold value as well as the deviation value and deviation ratio;
[0034] Step 44, when there is an abnormal power usage situation, determine the type of the abnormal power usage situation, the type of abnormal power usage situation includes: whether other devices are connected, whether a fault occurs, and whether the power usage data is maliciously tampered with;
[0035] Step 45: output corresponding alarm signals or control instructions according to the type of abnormal power usage.
[0036] Under the best implementation conditions, the beneficial effects produced are: it can effectively identify and handle abnormal power usage, improve power safety and reliability, and give corresponding alarm signals or control instructions, such as sound and light alarms, power off control, remote notification, etc.
[0037] In the best implementation, the technical problem to be solved is: how to effectively identify and classify the types of abnormal power usage and improve the accuracy and efficiency of classification.
[0038] In the best implementation, the technical solution adopted is:
[0039] Step 441, taking the deviation value and deviation ratio between the load power and the preset power as the original feature, and normalizing the original feature;
[0040] Step 442, based on the normalized original features, a feature extraction algorithm is used to extract the deviation direction and fluctuation features between the load power and the preset power as classification features;
[0041] Step 443, labeling the historical data according to the types of all abnormal power usage conditions as classification labels;
[0042] Step 444, training a classifier using the classification features and the classification labels, and evaluating its accuracy;
[0043] Step 445: Use the trained classifier to classify the deviation direction and fluctuation characteristics between the current load power and the preset power to obtain the type of abnormal power usage.
[0044] Under the best implementation conditions, the beneficial effects produced are: the types of abnormal power usage can be effectively identified and classified, the accuracy and efficiency of classification can be improved, and a basis can be provided for subsequent alarms or controls.
[0045] In the best implementation, the technical problem to be solved is: how to effectively analyze and optimize normal power consumption and give corresponding power consumption suggestions or optimization plans.
[0046] In the best implementation, the technical solution adopted is:
[0047] Step 46, when there is no abnormal power consumption, analyzing the deviation value and deviation ratio between the load power and the preset power, and providing corresponding power consumption suggestions or optimization solutions according to the analysis results;
[0048] Step 47, the abnormal power usage and power usage suggestions or optimization solutions are sent to the user's terminal device via wireless transmission, and displayed and interacted on the terminal device.
[0049] Under the best implementation conditions, the beneficial effects produced are: it can effectively analyze and optimize normal electricity usage, improve electricity efficiency and save electricity costs, and give corresponding electricity usage suggestions or optimization plans, which are sent to the user's terminal device via wireless transmission and displayed and interacted on the terminal device, making it convenient for users to understand and manage electricity usage.
[0050] In the best implementation, the technical problem to be solved is: how to effectively calculate and optimize the power efficiency and power cost of the load circuit and obtain the optimal power consumption plan.
[0051] In the best implementation, the technical solution adopted is:
[0052] Step 461, calculating the power efficiency and power cost of the load circuit according to the deviation value and deviation ratio between the load power and the preset power, and using them as indicators for power analysis;
[0053] Step 462, based on the power consumption analysis index, a data mining algorithm is used to perform cluster analysis on the power consumption pattern of the load circuit, and different power consumption categories are obtained;
[0054] Step 463, according to different power consumption categories, use an optimization algorithm to optimize and adjust the power consumption parameters of the load circuit, and obtain the optimal power consumption plan;
[0055] Step 464: Output the power consumption analysis indicators, power consumption categories and optimal power consumption plan as power consumption recommendations or optimization plans.
[0056] Under the best implementation conditions, the beneficial effects produced are: the power efficiency and power cost of the load circuit can be effectively calculated and optimized, and the optimal power consumption plan can be obtained, thereby improving power efficiency and saving power costs.
[0057] A smart meter electricity consumption monitoring system, comprising:
[0058] The data prediction module is used to collect historical data and current data in the smart meter database, use deep learning algorithms to predict the power of electrical equipment in the next hour, and use the predicted value as the preset power;
[0059] A signal conversion module is used to collect voltage signals and current signals of the load circuit and convert the voltage signals and current signals into digital signals;
[0060] A load power calculation module, used for calculating the load power according to the converted digital signal;
[0061] The abnormality detection module is used to compare the load power with the preset power, and determine whether there is abnormal power consumption based on the comparison result, and output corresponding alarm signals or control instructions based on the abnormal power consumption.
[0062] The beneficial effects produced are: it can realize the prediction of electricity demand and the power of electrical equipment in order to optimize the dispatch and management of the power grid; it can detect abnormal electricity consumption, such as electricity theft, failure, equipment replacement, etc., and timely alarm or control; it can analyze the electricity consumption pattern and efficiency, and give corresponding electricity consumption suggestions or optimization plans; it can send electricity consumption information to the user's terminal device through wireless transmission, and realize interaction with the user.
[0063] In a preferred implementation, the technical problem to be solved is: how to effectively receive and store the digital signal of the load loop to improve the calculation accuracy and stability of the load power.
[0064] In the best implementation, the technical solution adopted is:
[0065] The load power calculation module specifically includes:
[0066] The signal receiving submodule is used to receive digital signals and store them in two arrays, which respectively represent digital voltage signals and digital current signals;
[0067] The signal processing submodule is used to filter the digital signals in the two arrays to remove noise and interference and obtain smoothed digital signals;
[0068] The calculation submodule is used to calculate the load power according to the smoothed digital signal. The calculation formula is as follows:
[0069]
[0070] Where P is the load power, T is the cycle time, v(t) is the smoothed digital voltage signal, i(t) is the smoothed digital current signal, and t is the variable value of the time interval length in calculating the instantaneous power.
[0071] Under the best implementation conditions, the beneficial effects produced are: improving the calculation accuracy and stability of the load power, which can accurately reflect the power consumption of the load circuit; simplifying the structure and design of the load power calculation module, reducing the cost and maintenance difficulty; increasing the versatility and compatibility of the load power calculation module, which can adapt to load circuits of different types and specifications.
[0072] In the best implementation, the technical problem to be solved is: how to effectively decompose and encapsulate the anomaly detection module and improve its functionality and scalability.
[0073] In the best implementation, the technical solution adopted is:
[0074] The anomaly detection module specifically includes:
[0075] The standard setting submodule is used to set the judgment criteria and thresholds for abnormal power usage;
[0076] The power comparison submodule is used to compare the load power with the preset power to obtain a deviation value and a deviation ratio between the load power and the preset power;
[0077] The abnormality judgment submodule is used to judge whether there is abnormal power consumption according to the judgment standard and threshold value as well as the deviation value and deviation ratio;
[0078] The type judgment submodule is used to judge the type of abnormal power usage when there is abnormal power usage, and the type of abnormal power usage includes: whether other devices are connected, whether a fault occurs, and whether the power usage data is maliciously tampered with;
[0079] The alarm submodule is used to output corresponding alarm signals or control instructions according to the type of abnormal power consumption;
[0080] The power consumption plan acquisition module is used to analyze the deviation value and deviation ratio between the load power and the preset power when there is no abnormal power consumption, give corresponding power consumption suggestions or optimization plans according to the analysis results, and send the abnormal power consumption and power consumption suggestions or optimization plans to the user's terminal device via wireless transmission.
[0081] Under the best implementation conditions, the beneficial effects produced are: the functionality and scalability of the abnormality detection module are improved, and it can effectively judge and handle abnormal power usage of different types and degrees, and give corresponding alarm signals or control instructions. At the same time, in the absence of abnormal power usage, it provides users with power usage suggestions or optimization plans to improve power usage efficiency; the function and structure of the abnormality detection module are decomposed to improve its modularity and reusability; the interface and parameters of the abnormality detection module are encapsulated to improve its closedness and security. BRIEF DESCRIPTION OF THE DRAWINGS
[0082] Figure 1 is a flow chart of the method of the present invention;
[0083] Figure 2 is a flow chart of a preset power acquisition process of the present invention;
[0084] Figure 3 is a flow chart of a load power calculation process of the present invention;
[0085] Figure 4 It is a flow chart of the abnormal power consumption judgment and handling process of the present invention;
[0086] Figure 5 It is a schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION
[0087] In order to enable those skilled in the art to better understand the technical solution, the technical solution is described in detail below in conjunction with the embodiments. The description in this section is only exemplary and explanatory and should not have any limiting effect on the scope of protection of the present invention.
[0088] Embodiment 1:
[0089] This embodiment provides a smart meter power consumption monitoring method, which can monitor and analyze the power consumption of the load circuit in real time, and make corresponding alarms or controls according to the monitoring and analysis results to improve power consumption efficiency and safety. Figure 1 As shown, the method comprises the following steps:
[0090] Step 1: The data prediction module collects historical data and current data in the smart meter database, uses a deep learning algorithm to predict the power of the electrical equipment in the next hour, and uses the predicted value as the preset power;
[0091] The purpose of this step is to use the existing data in the smart meter database to establish a power prediction model for electrical equipment through a deep learning algorithm, so that the power of electrical equipment in the next hour can be predicted based on current and historical power consumption. The predicted value can be used as the preset power to facilitate subsequent comparison and judgment.
[0092] Specifically, Figure 2 As shown, this step includes the following sub-steps:
[0093] Step 11, collecting historical data and current data in the smart meter database;
[0094] This sub-step is to obtain the required data from the smart meter database, including historical data and current data. Historical data refers to the power data of the electrical equipment at each moment or each time interval (e.g., every minute) in a certain period of time in the past (e.g., one month, one year, etc.); current data refers to the power data of the electrical equipment at the most recent moment or the most recent time interval. These data can be automatically collected and stored in the database by the smart meter, or obtained through remote meter reading or other methods.
[0095] For example, in this embodiment, a smart meter is used, which can collect and record the power data of the electric devices connected to the load circuit every minute and store it in the database. Then, in this sub-step, the power data of the electric devices every minute in the past month can be collected as historical data, and the power data of the electric devices in the most recent minute can be collected as current data.
[0096] Step 12, preprocessing the collected data, including removing noise, outliers and missing values;
[0097] This sub-step is to perform some necessary preprocessing operations on the collected data to improve data quality and reliability and provide better input for subsequent deep learning model training. Preprocessing operations include removing noise, outliers, and missing values.
[0098] Noise removal refers to eliminating or reducing data deviations or fluctuations caused by factors such as acquisition equipment, transmission signals, and environmental interference. Noise removal can be achieved by filtering, smoothing, dimensionality reduction, and other methods.
[0099] Removing outliers means identifying, deleting or replacing data points that are obviously inconsistent with the normal range or distribution. Outliers may be caused by collection errors, failures, human interference, etc. Statistical analysis, cluster analysis, anomaly detection and other methods can be used to remove outliers.
[0100] Removing missing values refers to dealing with data gaps caused by collection failure, data loss, incomplete data, etc. Missing values may affect the integrity and continuity of the data. Missing values can be removed by deletion, interpolation, estimation and other methods.
[0101] For example, in this sub-step, the following preprocessing operations are performed on the collected power data of the electrical equipment:
[0102] A low-pass filter is used to remove high-frequency noise, making the data smoother;
[0103] A standard deviation-based anomaly detection algorithm is used to identify and remove data points that exceed the mean plus or minus three standard deviations to eliminate the impact of extreme values;
[0104] A linear interpolation-based method is used to fill in data gaps caused by acquisition failure or data loss to ensure data integrity and continuity.
[0105] Step 13, using the preprocessed data to train a deep learning model and evaluate the performance of the model;
[0106] This sub-step is to use the preprocessed data to establish a power prediction model for electrical equipment through a deep learning algorithm and evaluate the performance of the model. There are many types and structures of deep learning algorithms, such as convolutional neural networks, recurrent neural networks, long short-term memory networks, etc. According to different application scenarios and goals, appropriate deep learning algorithms and parameters can be selected for model training. In this embodiment, a long short-term memory network (LSTM) is used as a deep learning algorithm to establish a power prediction model for electrical equipment.
[0107] Step 14: Use the trained deep learning model to predict the power of the electrical equipment in the next hour, and use the predicted value as the preset power.
[0108] This sub-step uses the trained deep learning model to predict the power of electrical equipment at each time interval (e.g., every minute) for the current data and the next hour, and uses the predicted value as the preset power; this sub-step can be executed periodically or on demand, and the preset power is stored in the smart meter for subsequent comparison and judgment.
[0109] For example, in this sub-step, the trained LSTM network is used to predict the power data of the power consumption equipment in the last minute, and the power prediction value of the power consumption equipment every minute in the next hour is obtained. These preset powers represent the power expectation value of the power consumption equipment every minute in the next hour, which can reflect the power consumption law and change trend of the power consumption equipment.
[0110] Step 2: The signal conversion module collects the voltage signal and the current signal of the load circuit, and converts the voltage signal and the current signal into a digital signal;
[0111] The purpose of this step is to obtain the real-time power consumption of the load circuit, including voltage signals and current signals, and convert these signals into digital signals for subsequent calculation and processing.
[0112] Specifically, this step includes the following sub-steps:
[0113] Step 21, collecting voltage signals and current signals of the load circuit;
[0114] This sub-step is to collect the voltage signal and current signal of the electrical equipment connected to the load circuit through a smart meter or other collection equipment. These signals can reflect the working status and power consumption of the electrical equipment. This sub-step can be executed periodically or on demand, and the collected signals are stored in the smart meter or other collection equipment.
[0115] Step 22, converting the voltage signal and the current signal into digital signals;
[0116] This sub-step is to convert the collected analog signal into a digital signal to facilitate subsequent calculation and processing; this sub-step can be implemented by an analog-to-digital converter (ADC) or other conversion devices.
[0117] Specifically, this sub-step includes the following sub-steps:
[0118] Sampling an analog signal means obtaining a value of the analog signal at a certain time interval (e.g., every millisecond);
[0119] Quantize the analog signal, that is, map the value of the analog signal to an integer value within a certain range (e.g. 0-255);
[0120] Encode the analog signal, that is, represent the quantized integer value as a binary number or other base number.
[0121] Step 3, the load power calculation module calculates the load power according to the converted digital signal;
[0122] The purpose of this step is to calculate the real-time power of the electrical equipment connected to the load circuit according to the converted digital signal, and use the calculation result as the load power.
[0123] Specifically, Figure 3 As shown, this step includes the following sub-steps:
[0124] Step 31, receiving digital signals and storing them in two arrays, respectively representing digital voltage signals and digital current signals;
[0125] This sub-step is to receive the converted digital signal from the smart meter or other conversion device and store it in two arrays, representing the digital voltage signal and the digital current signal respectively.
[0126] For example, in this sub-step, a smart meter is used, which can receive and store the converted digital signals and store them in two arrays, representing the digital voltage signal and the digital current signal respectively. Then, in this sub-step, the following arrays can be obtained:
[0127] Digital voltage signal array: [172,171,173,…];
[0128] Digital current signal array: [43,44,42,…];
[0129] These arrays represent the digital voltage and current values of the power-consuming devices every millisecond.
[0130] Step 32, filtering the digital signals in the two arrays to remove noise and interference and obtain smoothed digital signals;
[0131] This sub-step is to filter the digital signals in the two arrays to remove the noise and interference that may be generated during the acquisition, conversion, transmission, etc., and obtain a smoothed digital signal. Filtering can improve data quality and reliability and provide more accurate input for subsequent power calculations; filtering can use filters of different types and parameters such as low-pass filters, high-pass filters, band-pass filters, and band-stop filters.
[0132] For example, in this sub-step, a low-pass filter is used to filter the digital signals in the two arrays and obtain smoothed digital signals. A low-pass filter is a filter that only allows low-frequency signals to pass through and suppresses high-frequency signals; a low-pass filter can eliminate or reduce data deviations or fluctuations caused by high-frequency noise or interference; the parameters of a low-pass filter include cutoff frequency, passband gain, stopband attenuation, etc. These parameters can be adjusted and optimized according to different data characteristics and requirements.
[0133] In this sub-step, a low-pass filter with a cutoff frequency of 50Hz, a passband gain of 1, and a stopband attenuation of 40dB is used to filter the digital signals in the two arrays and obtain the smoothed digital signals. Then, in this sub-step, the following arrays can be obtained:
[0134] The smoothed digital voltage signal array: [172,171.5,172.5,…];
[0135] Smoothed digital current signal array: [43,43.5,42.5,…];
[0136] These arrays represent the smoothed digital voltage and current values of the electrical equipment every millisecond after being processed by a low-pass filter.
[0137] Step 33, calculate the load power according to the smoothed digital signal, and the calculation formula is as follows:
[0138]
[0139] Where P is the load power, T is the cycle time, v(t) is the smoothed digital voltage signal, i(t) is the smoothed digital current signal, and t is the variable value of the time interval length in calculating the instantaneous power.
[0140] This sub-step is to calculate the real-time power of the electric device connected to the load circuit according to the smoothed digital signal, and use the calculation result as the load power. This sub-step can be implemented by a smart meter or other computing devices.
[0141] Specifically, this sub-step includes the following sub-steps:
[0142] According to the smoothed digital signal, the instantaneous power in each time interval (for example, each millisecond) is calculated, that is, v(t)i(t);
[0143] Based on the instantaneous power, calculate the average power in each cycle (for example, per second), that is,
[0144]
[0145] The average power is used as the load power and stored or sent to other terminal devices.
[0146] Step 4, compare the load power with the preset power, and determine whether there is abnormal power consumption according to the comparison result, and output a corresponding alarm signal or control instruction according to the abnormal power consumption.
[0147] The purpose of this step is to compare the real-time power of the electrical equipment connected to the load circuit with the predicted preset power, and determine whether there is abnormal power consumption based on the comparison result, and output corresponding alarm signals or control instructions based on the abnormal power consumption.
[0148] Specifically, Figure 4 As shown, this step includes the following sub-steps:
[0149] Step 41, setting the judgment criteria and thresholds for abnormal power usage;
[0150] This sub-step is to set the judgment criteria and thresholds for abnormal power usage to facilitate subsequent comparison and judgment. The judgment criteria refer to indicators used to judge whether there is abnormal power usage, such as the deviation value or deviation ratio between the load power and the preset power; the threshold refers to the critical value used to judge whether there is abnormal power usage, such as the deviation value or deviation ratio exceeding a certain value before abnormal power usage is considered. The judgment criteria and thresholds can be set and adjusted according to different power-consuming equipment and needs.
[0151] For example, in this sub-step, the following judgment criteria and thresholds are set:
[0152] Judgment standard: the deviation ratio between load power and preset power, i.e. (P-preset power) / preset power;
[0153] Threshold: ±10%, that is, when the deviation ratio is greater than 10% or less than -10%, it is considered that there is abnormal power consumption.
[0154] These judgment criteria and thresholds indicate that when the deviation ratio between the load power and the preset power is greater than 10% or less than -10%, it is considered that there is abnormal power consumption and an alarm or control is required.
[0155] Step 42, comparing the load power with the preset power to obtain a deviation value and a deviation ratio between the load power and the preset power;
[0156] This sub-step is to compare the real-time power of the electrical equipment connected to the load circuit with the predicted preset power, and obtain the deviation value and deviation ratio between the load power and the preset power. These deviation values and deviation ratios can reflect the degree of difference between the actual power consumption of the electrical equipment and the expected power consumption.
[0157] Step 43, judging whether there is abnormal power consumption according to the judgment standard and threshold value as well as the deviation value and deviation ratio;
[0158] This sub-step is to determine whether there is abnormal power usage based on the set judgment criteria and thresholds as well as the calculated deviation value and deviation ratio. If there is abnormal power usage, subsequent type judgment, alarm or control is required; if there is no abnormal power usage, subsequent power consumption analysis, suggestions or optimization are required.
[0159] Specifically, this sub-step includes the following sub-steps:
[0160] According to the judgment standard and threshold, the deviation value or deviation ratio in each time interval (for example, every minute) is judged. If it exceeds the threshold, it is considered that there is abnormal power consumption, otherwise it is considered that there is no abnormal power consumption;
[0161] The judgment result is stored or sent to other terminal devices.
[0162] Step 44, when there is an abnormal power usage situation, determine the type of the abnormal power usage situation, the type of abnormal power usage situation includes: whether other devices are connected, whether a fault occurs, and whether the power usage data is maliciously tampered with;
[0163] This sub-step is to determine the type of abnormal power usage when there is abnormal power usage, so as to facilitate subsequent alarm or control. The types of abnormal power usage include: whether other devices are connected, whether a fault occurs, and whether the power usage data is maliciously tampered with. These types can reflect the cause and nature of the abnormal power usage.
[0164] Specifically, this sub-step includes the following sub-steps:
[0165] Step 441, taking the deviation value and deviation ratio between the load power and the preset power as the original feature, and normalizing the original feature;
[0166] This sub-step is to take the deviation value and deviation ratio between the load power and the predicted preset power as the original features, and normalize the original features; normalization refers to mapping the value of the original feature to a fixed range (for example, 0-1) to eliminate the dimension and scale differences between different features and improve the stability and comparability of the data. Normalization can adopt maximum and minimum normalization, mean variance normalization, linear normalization and other methods.
[0167] Step 442, based on the normalized original features, a feature extraction algorithm is used to extract the deviation direction and fluctuation features between the load power and the preset power as classification features;
[0168] This sub-step is to extract the deviation direction and fluctuation characteristics between the load power and the preset power as classification features based on the normalized original features using a feature extraction algorithm; the feature extraction algorithm refers to extracting more representative and distinctive new features from the original features to facilitate subsequent classification; the deviation direction and fluctuation characteristics between the load power and the preset power refer to new features that reflect the changing trend and magnitude of the load power and the preset power, and these new features can better describe the type of abnormal power consumption.
[0169] Specifically, this sub-step includes the following sub-steps:
[0170] According to the normalized deviation ratio, the deviation direction in each time interval (e.g., every minute) is calculated, that is, whether the load power increases or decreases relative to the preset power;
[0171] According to the normalized deviation ratio, the deviation fluctuation within each time interval (for example, every minute) is calculated, that is, the variation range of the load power relative to the preset power;
[0172] The deviation direction and deviation fluctuation are used as classification features and stored or sent to other terminal devices.
[0173] Step 443, labeling the historical data according to the types of all abnormal power usage conditions as classification labels;
[0174] This sub-step is to mark the historical data according to the types of all abnormal power usage as classification labels; classification labels refer to symbols or numbers that represent the types of abnormal power usage, for example, 1 indicates that other equipment is connected, 2 indicates a fault, 3 indicates whether the power usage data has been maliciously tampered with, etc. Classification labels can be used for subsequent classifier training and evaluation.
[0175] Specifically, this sub-step includes the following sub-steps:
[0176] Define a classification label set based on the types of all abnormal power consumption situations, such as {1, 2, 3};
[0177] According to the actual situation in the historical data, the data in each time interval (for example, every minute) is labeled, that is, a classification label is assigned to each data;
[0178] The annotated data is stored or sent to other terminal devices.
[0179] Step 444, training a classifier using a classification algorithm based on the labeled data, and evaluating the performance of the classifier;
[0180] Specifically, this sub-step includes the following sub-steps:
[0181] According to the labeled data, select the appropriate classification algorithm and parameters for classifier training; there are many types and structures of classification algorithms, such as decision tree, support vector machine, naive Bayes, neural network, etc. According to different data characteristics and requirements, you can choose the appropriate classification algorithm and parameters for classifier training.
[0182] The labeled data is divided into three parts: training set, validation set and test set. The training set is the data used to actually train the classifier; the validation set is the data used to test the classifier performance and adjust hyperparameters (such as learning rate, regularization coefficient, etc.) during the training process; the test set is the data used to evaluate the final performance of the classifier after training is completed and avoid overfitting (that is, the classifier performs well on the training data, but performs poorly on new data).
[0183] After each round of training, the indicators on the validation set are used to test the performance of the classifier, and whether to continue training or adjust the hyperparameters is determined based on whether the indicators on the validation set are improved. After the training is completed, the indicators on the test set are used to evaluate the final performance of the classifier and compare it with other classifiers or benchmarks.
[0184] Step 445, using the trained classifier to classify the deviation direction and fluctuation characteristics between the current load power and the preset power, and obtain the type of abnormal power consumption;
[0185] This sub-step uses a trained classifier to classify the deviation direction and fluctuation characteristics between the current load power and the preset power, and obtains the type of abnormal power consumption. This sub-step can be executed periodically or on demand, and the classification results can be stored or sent to other terminal devices.
[0186] Specifically, this sub-step includes the following sub-steps:
[0187] Receive the current deviation direction and deviation fluctuation as input features and input them into the classifier;
[0188] According to the probability distribution of the classifier output, the category with the highest probability is selected as the output category, that is, the type of abnormal power consumption;
[0189] Store or send output categories to other terminal devices.
[0190] Step 45, outputting corresponding alarm signals or control instructions according to the type of abnormal power consumption;
[0191] This sub-step is to output a corresponding alarm signal or control instruction according to the type of abnormal power usage; an alarm signal refers to a signal that prompts the user or manager of the existence of abnormal power usage through sound, light, text, etc.; a control instruction refers to an instruction to handle the abnormal power usage by disconnecting or adjusting the power supply or power of the electrical equipment connected to the load circuit.
[0192] Specifically, this sub-step includes the following sub-steps:
[0193] According to the type of abnormal power consumption, select the appropriate alarm signal or control command, which can be set and adjusted according to the severity and handling method of different abnormal power consumption situations;
[0194] Output alarm signals or control instructions to the user's or manager's terminal device, or directly control the electrical equipment connected to the load circuit.
[0195] Step 46, when there is no abnormal power consumption, analyzing the deviation value and deviation ratio between the load power and the preset power, and providing corresponding power consumption suggestions or optimization solutions according to the analysis results;
[0196] This step is to analyze the deviation value and deviation ratio between the load power and the preset power when there is no abnormal power consumption, and give corresponding power consumption suggestions or optimization plans based on the analysis results.
[0197] Specifically, this step includes the following sub-steps:
[0198] Step 461, calculating the power efficiency and power cost of the load circuit according to the deviation value and deviation ratio between the load power and the preset power, and using them as indicators for power analysis;
[0199] This sub-step is to calculate the power efficiency and power cost of the load circuit according to the deviation value and deviation ratio between the load power and the preset power, and use them as indicators for power analysis. Power efficiency refers to the ratio of the output power to the input power of the power equipment connected to the load circuit, reflecting the energy conversion efficiency of the power equipment; power cost refers to the product of the power consumed by the power equipment connected to the load circuit within a certain period of time and the electricity price, reflecting the economy of the power equipment. These indicators can evaluate the power performance and level of the load circuit.
[0200] Specifically, this sub-step includes the following sub-steps:
[0201] According to the load power and the preset power, the power efficiency in each time interval (for example, every minute) is calculated, that is, load power / preset power;
[0202] According to the load power and the electricity price, the electricity cost in each time interval (for example, every minute) is calculated, that is, the load power time electricity price;
[0203] The electricity efficiency and electricity cost are stored or sent to other terminal devices.
[0204] Step 462, based on the power consumption analysis index, a data mining algorithm is used to perform cluster analysis on the power consumption pattern of the load circuit, and different power consumption categories are obtained;
[0205] This sub-step uses a data mining algorithm to cluster the power consumption patterns of the load circuits according to the power consumption analysis indicators, and obtains different power consumption categories. Data mining algorithms refer to algorithms that extract valuable information and knowledge from large amounts of data; cluster analysis refers to a data mining method that groups data according to metrics such as similarity or distance; power consumption categories refer to the classification of the power consumption patterns of the load circuits into different types, such as high efficiency, energy saving, waste, etc., according to the power consumption analysis indicators. These power consumption categories can reflect the power consumption characteristics and rules of the load circuits.
[0206] Specifically, this sub-step includes the following sub-steps:
[0207] receiving the indicators of electricity consumption analysis, namely, electricity consumption efficiency and electricity consumption cost, and using them as inputs of the data mining algorithm;
[0208] Select appropriate data mining algorithms and parameters for cluster analysis, such as K-means algorithm, hierarchical clustering algorithm, etc., and determine the number and center of clusters;
[0209] According to the results of cluster analysis, each cluster is assigned a label of electricity consumption category and output to other terminal devices.
[0210] Step 463, according to different power consumption categories, an optimization algorithm is used to optimize and adjust the power consumption parameters of the load circuit, and the optimal power consumption plan is obtained.
[0211] Specifically, this sub-step includes the following sub-steps:
[0212] receiving the electricity usage category and using it as input to the optimization algorithm;
[0213] Select appropriate optimization algorithms and parameters for optimization, such as genetic algorithms, simulated annealing algorithms, etc., and determine the optimization objective function and constraints;
[0214] Based on the results of the optimization and adjustment, an optimal power consumption plan is generated for each power consumption category and output to other terminal devices.
[0215] Step 464: Output the power consumption analysis indicators, power consumption categories and optimal power consumption plan as power consumption recommendations or optimization plans.
[0216] Step 47, the abnormal power usage and power usage suggestions or optimization solutions are sent to the user's terminal device via wireless transmission, and displayed and interacted on the terminal device.
[0217] Specifically, this step includes the following sub-steps:
[0218] Pack the abnormal power usage and power usage suggestions or optimization plans into a data packet through a computer, and encrypt and compress it;
[0219] The computer selects a suitable wireless transmission protocol and channel for wireless transmission, such as Wi-Fi Direct, Bluetooth Low Energy, etc., and determines the target address and port for transmission;
[0220] Receive, decrypt and decompress data packets on the user's terminal device and store the data locally or in the cloud;
[0221] Design and implement a graphical interface on the user's terminal device to display and interact with data and provide corresponding operation functions.
[0222] Embodiment 2:
[0223] In this embodiment, a smart meter electricity consumption monitoring system is used. Figure 5 As shown, the system includes a data prediction module, a signal conversion module, a load power calculation module and an anomaly detection module. The working process of the system is as follows:
[0224] The data prediction module collects historical data and current data in the smart meter database, uses a deep learning algorithm to predict the power of electrical equipment in the next hour, and uses the predicted value as the preset power;
[0225] The signal conversion module collects the voltage signal and current signal of the load circuit, and converts the voltage signal and current signal into digital signals;
[0226] The load power calculation module calculates the load power according to the converted digital signal;
[0227] The abnormality detection module compares the load power with the preset power, and determines whether there is abnormal power consumption according to the comparison result, and outputs corresponding alarm signals or control instructions according to the abnormal power consumption;
[0228] When there is no abnormal power consumption, the abnormal detection module analyzes the deviation value and deviation ratio between the load power and the preset power, and gives corresponding power consumption suggestions or optimization plans based on the analysis results;
[0229] The abnormality detection module sends abnormal power usage and power usage suggestions or optimization plans to the user terminal device via wireless transmission, and displays and interacts on the terminal device.
[0230] In this example, the following data is used:
[0231] There are three electrical devices in the factory, namely air conditioner, refrigerator and water heater, which are connected to the same load circuit;
[0232] The smart meter database stores the power data of the three electrical devices every minute in the past year, as well as the power data of the current time (10:00 on October 25, 2023);
[0233] The data prediction module uses a deep learning algorithm based on the long short-term memory network (LSTM) to predict the power of electrical equipment in the next hour (i.e. 10:00-11:00) and uses the predicted value as the preset power;
[0234] The signal conversion module collects the voltage signal and current signal of the load circuit and converts them into digital signals with a sampling frequency of 1000Hz;
[0235] The load power calculation module calculates the load power according to the converted digital signal. The calculation formula is as follows:
[0236]
[0237] Where P is the load power, T is the cycle time, v(t) is the smoothed digital voltage signal, i(t) is the smoothed digital current signal, and t is the variable value of the time interval length in calculating the instantaneous power.
[0238] The abnormal detection module sets the judgment criteria and thresholds for abnormal power usage, that is, when the deviation between the load power and the preset power is greater than 100W or the deviation ratio is greater than 10%, it is considered that there is abnormal power usage;
[0239] The anomaly detection module uses a feature extraction and classification algorithm based on support vector machine (SVM) to judge the type of abnormal power usage, including whether other devices are connected, whether there is a fault, and whether the power usage data has been maliciously tampered with;
[0240] The abnormality detection module outputs corresponding alarm signals or control instructions according to the type of abnormal power consumption, such as sound and light alarm, disconnection of load circuit, and SMS notification.
[0241] When there is no abnormal power consumption, the abnormal detection module analyzes the deviation value and deviation ratio between the load power and the preset power, and gives corresponding power consumption suggestions or optimization plans based on the analysis results, such as adjusting the air conditioning temperature, cleaning the refrigerator fan, and timing the water heater off.
[0242] The abnormality detection module sends abnormal power usage and power usage suggestions or optimization solutions to the user terminal device through wireless transmission, and displays and interacts on the terminal device. The user terminal device can be a smartphone, tablet computer, laptop computer, etc.
[0243] The following uses the data from 10:00 to 10:05 as an example to illustrate the working process of the system:
[0244] The data prediction module collects the power data of the three electrical devices every minute in the past year, as well as the power data of the current time (10:00 on October 25, 2023). The data prediction module preprocesses the collected data, including removing noise, outliers, and missing values. The data prediction module uses the preprocessed data to train a LSTM-based deep learning model and evaluates the performance of the model. The data prediction module uses the trained deep learning model to predict the power of the electrical equipment in the next hour (i.e., 10:00-11:00) and uses the predicted value as the preset power. In this embodiment, the following data is used:
[0245]
[0246]
[0247] These data represent the power and preset power of electrical equipment per minute, where the preset power refers to the value obtained by predicting the power of electrical equipment in the next hour through a deep learning algorithm.
[0248] The signal conversion module collects the voltage signal and current signal of the load circuit and converts them into digital signals with a sampling frequency of 1000Hz.
[0249] The load power calculation module receives the digital signal and stores it in two arrays, which represent the digital voltage signal and the digital current signal respectively;
[0250] The load power calculation module filters the digital signals in the two arrays to remove noise and interference and obtain smoothed digital signals. The load power calculation module uses a wavelet transform-based filtering algorithm to perform multi-scale decomposition and reconstruction of digital signals, thereby retaining the main features of the signal and eliminating noise and interference.
[0251] The load power calculation module calculates the load power based on the smoothed digital signal. The calculation formula is as follows:
[0252]
[0253] Where P is the load power, T is the cycle time, v(t) is the smoothed digital voltage signal, i(t) is the smoothed digital current signal, and t is the variable value of the time interval length in calculating instantaneous power. The load power calculation module multiplies and sums the smoothed digital voltage signal and digital current signal in each cycle to obtain the load power in that cycle.
[0254] The abnormal detection module compares the load power with the preset power, and determines whether there is abnormal power usage based on the comparison result, and outputs corresponding alarm signals or control instructions based on the abnormal power usage. The abnormal detection module sets the judgment criteria and thresholds for abnormal power usage, that is, when the deviation value between the load power and the preset power is greater than 100W or the deviation ratio is greater than 10%, it is considered that there is abnormal power usage. The abnormal detection module uses a feature extraction and classification algorithm based on SVM to judge the type of abnormal power usage, which includes: whether other devices are connected, whether a fault occurs, and whether the power usage data is maliciously tampered with. According to the type of abnormal power usage, the abnormal detection module outputs corresponding alarm signals or control instructions, such as sound and light alarms, disconnecting the load circuit, sending SMS notifications, etc.
[0255] Abnormal power consumption is determined by whether the deviation value and deviation ratio exceed the threshold; the abnormal power consumption type is classified by the deviation direction and fluctuation characteristics between the load power and the preset power according to the SVM algorithm; the alarm signal or control instruction is output in an appropriate manner according to the abnormal power consumption type. For example, when the load power suddenly exceeds the preset power, resulting in abnormal power consumption, it is determined that other devices may be connected to the load circuit, and an audible and visual alarm and a text message are output to notify the user to check the load circuit.
[0256] When there is no abnormal power consumption, the abnormal detection module analyzes the deviation value and deviation ratio between the load power and the preset power, and gives corresponding power consumption suggestions or optimization plans based on the analysis results. The abnormal detection module calculates the power consumption efficiency and power consumption cost of the load circuit based on the deviation value and deviation ratio between the load power and the preset power, and uses them as indicators for power consumption analysis. The abnormal detection module uses a data mining algorithm to cluster the power consumption patterns of the load circuit based on the indicators of power consumption analysis, and obtains different power consumption categories. The abnormal detection module uses an optimization algorithm to optimize and adjust the power consumption parameters of the load circuit based on different power consumption categories, and obtains the optimal power consumption plan. The abnormal detection module uses the indicators of power consumption analysis, power consumption categories and the optimal power consumption plan as power consumption suggestions or optimization plans, and outputs them to the user.
[0257] The anomaly detection module sends abnormal power usage and power usage suggestions or optimization plans to the user terminal device through wireless transmission, and displays and interacts on the terminal device. The user terminal device can be a smartphone, tablet computer, laptop, etc. The anomaly detection module packages this information into a data packet, encrypts and compresses it. The anomaly detection module selects the appropriate wireless transmission protocol and channel for wireless transmission, such as Wi-Fi Direct, Bluetooth LowEnergy, etc., and determines the target address and port for transmission. The data packet is received, decrypted and decompressed on the user terminal device, and the data is stored locally or in the cloud.
[0258] It should be noted that, in this article, the terms "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. Specific examples are used in this article to illustrate the principle and implementation of the technical solution of the present invention. The above examples are only used to help understand the method of the present invention and its core idea. The above is only a preferred implementation of the present invention. It should be pointed out that due to the limitations of textual expression and the objective existence of infinite specific structures, for ordinary technicians in this technical field, without departing from the principle of the present invention, several improvements, modifications or changes can be made, and the above technical features can also be combined in an appropriate manner; these improvements, modifications, changes or combinations, or the direct application of the concept and technical solution of the invention to other occasions without improvement, should be regarded as the protection scope of the present invention.
Claims
1. A method for monitoring electricity consumption of a smart meter, characterized in that: The following steps are involved: Step 1: The data prediction module collects historical data and current data in the smart meter database, uses a deep learning algorithm to predict the power of the electrical equipment in the next hour, and uses the predicted value as the preset power; Step 2: The signal conversion module collects the voltage signal and the current signal of the load circuit, and converts the voltage signal and the current signal into a digital signal; Step 3, the load power calculation module calculates the load power according to the converted digital signal; Step 4: The abnormality detection module compares the load power with the preset power, and determines whether there is abnormal power consumption according to the comparison result, and outputs a corresponding alarm signal or control instruction according to the abnormal power consumption; The step 4 specifically includes: Step 41, setting the judgment criteria and thresholds for abnormal power usage; Step 42, comparing the load power with the preset power to obtain a deviation value and a deviation ratio between the load power and the preset power; Step 43, judging whether there is abnormal power consumption according to the judgment standard and threshold value as well as the deviation value and deviation ratio; Step 44, when there is an abnormal power usage situation, determine the type of the abnormal power usage situation, the type of abnormal power usage situation includes: whether other devices are connected, whether a fault occurs, and whether the power usage data is maliciously tampered with; Step 45, outputting corresponding alarm signals or control instructions according to the type of abnormal power consumption; The determination of the type of abnormal power usage specifically includes: Step 441, taking the deviation value and deviation ratio between the load power and the preset power as the original feature, and normalizing the original feature; Step 442, based on the normalized original features, a feature extraction algorithm is used to extract the deviation direction and fluctuation features between the load power and the preset power as classification features; Step 443, labeling the historical data according to the types of all abnormal power usage conditions as classification labels; Step 444, training a classifier using the classification features and the classification labels, and evaluating its accuracy; Step 445: Use the trained classifier to classify the current data to obtain the type of abnormal power usage.
2. The smart meter electricity consumption monitoring method according to claim 1, characterized in that: The step 1 specifically includes: Step 11, collecting historical data and current data in the smart meter database; Step 12, preprocessing the collected data, including removing noise, outliers and missing values; Step 13, using the preprocessed data to train a deep learning model and evaluate the performance of the model; Step 14: Use the trained deep learning model to predict the power of the electrical equipment in the next hour, and use the predicted value as the preset power.
3. The method for monitoring electricity consumption of a smart meter according to claim 1, characterized in that: The step 3 specifically includes: Step 31, receiving digital signals and storing them in two arrays, respectively representing digital voltage signals and digital current signals; Step 32, filtering the digital signals in the two arrays to remove noise and interference and obtain smoothed digital signals; Step 33, calculate the load power according to the smoothed digital signal, and the calculation formula is as follows: Wherein, P is the load power, T is the cycle time, v(t) is the smoothed digital voltage signal, and i(t) is the smoothed digital current signal.
4. The method for monitoring electricity consumption of a smart meter according to claim 1, characterized in that: The step 4 further comprises: Step 46, when there is no abnormal power consumption, analyzing the deviation value and deviation ratio between the load power and the preset power, and providing corresponding power consumption suggestions or optimization solutions according to the analysis results; Step 47, abnormal power usage, power usage suggestions or optimization solutions and related information are sent to the user's terminal device via wireless transmission, and displayed and interacted on the terminal device.
5. The method for monitoring electricity consumption of a smart meter according to claim 4, characterized in that: The step 46 specifically includes: Step 461, calculating the power efficiency and power cost of the load circuit according to the deviation value and deviation ratio between the load power and the preset power, and using them as indicators for power analysis; Step 462, based on the power consumption analysis index, a data mining algorithm is used to perform cluster analysis on the power consumption pattern of the load circuit, and different power consumption categories are obtained; Step 463, according to different power consumption categories, use an optimization algorithm to optimize and adjust the power consumption parameters of the load circuit, and obtain the optimal power consumption plan; Step 464: Output the power consumption analysis indicators, power consumption categories and optimal power consumption plan as power consumption recommendations or optimization plans.
6. A smart meter electricity consumption monitoring system based on the method described in any one of claims 1 to 5, characterized in that: include: The data prediction module is used to collect historical data and current data in the smart meter database, use deep learning algorithms to predict the power of electrical equipment in the next hour, and use the predicted value as the preset power; A signal conversion module is used to collect voltage signals and current signals of the load circuit and convert the voltage signals and current signals into digital signals; A load power calculation module, used for calculating the load power according to the converted digital signal; The abnormality detection module is used to compare the load power with the preset power, and determine whether there is abnormal power consumption based on the comparison result, and output corresponding alarm signals or control instructions based on the abnormal power consumption.
7. The smart meter electricity consumption monitoring system according to claim 6, characterized in that: The load power calculation module specifically includes: The signal receiving submodule is used to receive digital signals and store them in two arrays, which respectively represent digital voltage signals and digital current signals; The signal processing submodule is used to filter the digital signals in the two arrays to remove noise and interference and obtain smoothed digital signals; The calculation submodule is used to calculate the load power according to the smoothed digital signal. The calculation formula is as follows: Wherein, P is the load power, T is the cycle time, v(t) is the smoothed digital voltage signal, and i(t) is the smoothed digital current signal.
8. The smart meter electricity consumption monitoring system according to claim 6, characterized in that: The anomaly detection module specifically includes: The standard setting submodule is used to set the judgment criteria and thresholds for abnormal power usage; The power comparison submodule is used to compare the load power with the preset power to obtain a deviation value and a deviation ratio between the load power and the preset power; The abnormality judgment submodule is used to judge whether there is abnormal power consumption according to the judgment standard and threshold value as well as the deviation value and deviation ratio; The type judgment submodule is used to judge the type of abnormal power usage when there is abnormal power usage, and the type of abnormal power usage includes: whether other devices are connected, whether a fault occurs, and whether the power usage data is maliciously tampered with; The alarm submodule is used to output corresponding alarm signals or control instructions according to the type of abnormal power consumption; The power consumption plan acquisition module is used to analyze the deviation value and deviation ratio between the load power and the preset power when there is no abnormal power consumption, give corresponding power consumption suggestions or optimization plans according to the analysis results, and send the abnormal power consumption, power consumption suggestions or optimization plans and related information to the user's terminal device via wireless transmission.
