Intelligent overload early warning method and system for charging extension socket

Abstract

The invention provides an intelligent overload early warning method and system for a charging extension socket. The method comprises the following steps: synchronously sampling current and voltage of each socket, converting the current and voltage into digital signals, and obtaining an electricity consumption sequence of each socket; calculating the load proportion of each socket, and if the proportion exceeds a first preset threshold value, judging that abnormal distribution exists and extracting abnormal distribution characteristics; generating an abnormal oscillation feature set based on the abnormal distribution features; classifying the abnormal oscillation feature set, calculating an abnormal probability in combination with historical statistical data, and determining an abnormal oscillation state if the probability is higher than a second preset threshold value; and a decision tree algorithm is adopted to obtain a correlation intensity value of the abnormal oscillation state and the load proportion, incremental detection analysis is executed, and if an existing value in an incremental change sequence exceeds a third preset threshold value, an overload early warning signal is generated. According to the invention, through accurate load monitoring, efficient abnormal oscillation identification and dynamic correlation analysis, intelligent overload early warning is realized, and the safety of the charging extension socket is improved.

Description

Technical Field

[0001] The present invention relates to the field of power equipment safety technology, and in particular to an intelligent overload warning method and system for a charging socket. Background Art

[0002] The field of intelligent power management plays a vital role in modern energy systems and home safety. Its core lies in achieving precise monitoring and safety assurance of electrical equipment through technical means to improve energy efficiency and reduce the risk of electrical accidents. With the popularization of smart home and Internet of Things technologies, charging strips, as key nodes in power management, have a direct impact on the reliability and safety of the entire system due to their intelligence level. However, existing intelligent management methods for charging strips have significant limitations in practical applications. These limitations mainly include insufficient dynamic monitoring capabilities of power loads, difficulty in accurately distinguishing between normal fluctuations and abnormal conditions, and a lack of detailed analysis of the rationality of power distribution across multiple outlets. These shortcomings make the system prone to misjudgment or omission when faced with complex power usage scenarios, reducing security and user experience.

[0003] In this context, the core challenges facing research focus on how to achieve real-time and accurate acquisition of electricity consumption, dynamic calculation of load ratios, and accurate identification of abnormal oscillations. First, real-time acquisition of electricity consumption needs to overcome the accuracy and speed limitations of data acquisition to ensure the reliability of monitoring results. Second, the dynamic calculation of load ratios involves the analysis of electricity distribution among multiple sockets, and the problem of identifying uneven distribution or abnormal surges must be solved. Finally, the identification of abnormal oscillations requires the establishment of an accurate feature library to distinguish normal load fluctuations from potential danger signals, a process that is easily affected by the complexity of electricity usage scenarios and equipment differences. These unresolved technical factors make it difficult for the system to achieve efficient early warning and reasonable distribution in a changing environment, which in turn leads to safety hazards and reduced energy efficiency.

[0004] Therefore, how to design a method to obtain the power consumption of each socket in real time, dynamically calculate the load ratio, build an oscillation feature library to accurately distinguish normal and abnormal states, and timely warn of abnormal increments based on power distribution analysis has become a key issue in the research of smart charging socket overload warning. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent overload warning method for a charging socket, the method comprising: S101. Synchronously sampling the current and voltage at each outlet using a high-frequency sensor to obtain an analog signal, converting the analog signal into a digital signal using analog-to-digital conversion technology, denoising the digital signal, and obtaining a power consumption sequence for each outlet based on the denoised data. S102. Based on the power usage sequence of each socket, calculate the load ratio of each socket relative to the total load using a weighted average algorithm. If the load ratio is greater than a first preset threshold, determine that the socket corresponding to the load ratio has an abnormal distribution, and extract the abnormal distribution characteristics of the socket corresponding to the abnormal distribution; S103. Extracting the time series change rate from the abnormal distribution features to obtain a change rate sequence; obtaining a frequency component set based on the change rate sequence; separating high-frequency components from the frequency component set; calculating the oscillation amplitude of each high-frequency component to obtain an oscillation amplitude set; and, for each of the oscillation amplitude sets, marking an oscillation signal as an abnormal oscillation signal if the oscillation amplitude exceeds a preset threshold; and generating an abnormal oscillation feature set based on the abnormal oscillation signal. S104. Classify the abnormal oscillation feature set to obtain a classification result; based on the classification result and historical statistical data, calculate the probability value of the current oscillation being abnormal; if the probability value is higher than a second preset threshold, determine that the current oscillation is in an abnormal oscillation state; S105. Use a decision tree algorithm to obtain the correlation strength value between the abnormal oscillation state and the load ratio, perform incremental detection analysis on the correlation strength value, and obtain an incremental change sequence. If there is a value in the incremental change sequence that exceeds a third preset threshold, an overload warning signal is generated.

[0006] Preferably, the step S102 includes: Determine the weight of each socket according to the historical usage frequency and rated power of each socket; Calculating the actual power consumption of the socket according to the weight and the total power consumption, and obtaining the load ratio of the socket relative to the total load based on the actual power consumption; Comparing the load ratio of each socket with a first preset threshold, and if the load ratio is greater than the threshold, determining that the socket has abnormal allocation; For the sockets with abnormal distribution, abnormal distribution features of the sockets are extracted.

[0007] Preferably, the step S103 includes: Extract the time series data of power consumption of abnormal sockets from the abnormal distribution characteristics, calculate the change rate between adjacent time points, and obtain the change rate series; Convert the rate of change sequence from the time domain to the frequency domain using a fast Fourier transform to obtain a set of frequency components; Filtering out components with frequencies greater than a preset threshold from the frequency component set to obtain a high-frequency component set; For each component in the high-frequency component set, calculating its oscillation amplitude to obtain an oscillation amplitude set; The oscillation amplitude set is traversed, and if an oscillation amplitude is greater than a preset oscillation threshold, it is marked as an abnormal oscillation signal, and all the abnormal oscillation signals are aggregated to generate an abnormal oscillation feature set.

[0008] Preferably, the step S104 includes: Extract key features from the abnormal oscillation feature set, classify the data using a machine learning method based on the key features, and calculate the abnormal probability corresponding to each classification result in combination with historical data statistics; Assign weights based on the importance of features and calculate weighted anomaly probability; The weighted abnormal probability is compared with a second preset threshold value, and if the probability value is higher than the second preset threshold value, the current oscillation is determined to be an abnormal oscillation state.

[0009] Preferably, the step S105 includes: A data set is formed by combining the historical load ratio and abnormal oscillation state of the socket to train a decision tree model, and a correlation strength value between the abnormal oscillation state and the load ratio of the current socket is obtained based on the decision tree model; For the correlation strength value, calculating a change trend of the strength value to obtain an incremental change sequence; If there is a value in the incremental change sequence that exceeds a third preset threshold, an overload warning signal is generated. In a second aspect, the present invention provides an intelligent overload warning system for a charging socket, the system comprising: The module acquires the power consumption sequence of each socket by synchronously sampling the current and voltage of each socket through a high-frequency sensor to obtain an analog signal, converts the analog signal into a digital signal using analog-to-digital conversion technology, denoises the digital signal, and obtains the power consumption sequence of each socket based on the denoised data; an abnormal distribution feature extraction module, which calculates the load ratio of each socket relative to the total load using a weighted average algorithm based on the power consumption sequence of each socket, and if the load ratio is greater than a first preset threshold, determines that the socket corresponding to the load ratio has abnormal distribution, and extracts abnormal distribution features of the socket corresponding to the abnormal distribution; an abnormal oscillation feature set generation module, which extracts the time series change rate from the abnormal distribution feature to obtain a change rate sequence, obtains a frequency component set based on the change rate sequence, separates high-frequency components from the frequency component set, calculates the oscillation amplitude of each high-frequency component to obtain an oscillation amplitude set, and for the oscillation amplitude set, if an oscillation amplitude exceeds a preset threshold, marks it as an abnormal oscillation signal, and generates an abnormal oscillation feature set based on the abnormal oscillation signal; an abnormal oscillation state determination module, which classifies the abnormal oscillation feature set to obtain a classification result; calculates a probability value of the current oscillation being abnormal based on the classification result and historical statistical data; and determines that the current oscillation is in an abnormal oscillation state if the probability value is higher than a second preset threshold; The overload warning signal generation module adopts a decision tree algorithm to obtain the correlation strength value between the abnormal oscillation state and the load ratio, performs incremental detection analysis on the correlation strength value, and obtains an incremental change sequence. If there is a value in the incremental change sequence that exceeds a third preset threshold, an overload warning signal is generated.

[0010] In a third aspect, the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the intelligent overload warning method for the charging strip as described above is implemented.

[0011] The present invention provides an intelligent overload warning method for charging socket strips. By real-time monitoring of the load status, oscillation characteristics, and load distribution of each socket, combined with data analysis, signal processing, and machine learning technologies, it can accurately predict overload risks and generate warning signals to effectively ensure power safety.

[0012] Specific benefits include: High-frequency sensors synchronously sample the current and voltage at each outlet and convert the analog signals into digital signals using analog-to-digital conversion technology, ensuring real-time and accurate data acquisition. By denoising the digital signals, the power consumption series of each outlet is obtained and the load ratio is calculated using a weighted average algorithm, enabling accurate identification of load distribution anomalies and abnormal oscillation signals. A weighted average algorithm is used to calculate the load ratio of each outlet relative to the total load, and abnormal distribution is determined based on a preset threshold, enabling timely detection of uneven load distribution. The time series change rate is extracted from abnormal distribution features, and high-frequency components are separated using a fast Fourier transform to calculate the oscillation amplitude, enabling accurate identification of abnormal oscillation signals. A decision tree algorithm is used to analyze the correlation strength between abnormal oscillation states and load ratios, and incremental detection is used to analyze load change trends, enabling early warning of potential overload risks. Classifying the abnormal oscillation feature set and calculating the anomaly probability based on historical statistical data improves the accuracy and reliability of early warnings. Through real-time monitoring and precise analysis, overload risks can be detected and warned in advance, preventing safety incidents such as equipment damage and fires caused by overload. By identifying abnormal distribution, users can be guided to distribute loads appropriately, thus extending the service life of charging strips. The method and system of the present invention can be widely applied to various types of charging strips and power distribution equipment, and have high practicality and promotion value. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1This is a flow chart of an intelligent overload warning method for a charging socket strip according to the present invention.

[0014] Figure 2 Detailed flowchart of step S102 of the present invention.

[0015] Figure 3 Detailed flowchart of step S103 of the present invention. DETAILED DESCRIPTION

[0016] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0017] An embodiment of the present invention provides an intelligent overload warning method for a charging socket, the method comprising: S101. Synchronously sampling the current and voltage at each outlet using a high-frequency sensor to obtain an analog signal, converting the analog signal into a digital signal using analog-to-digital conversion technology, denoising the digital signal, and obtaining a power consumption sequence for each outlet based on the denoised data. In this embodiment, a high-frequency sensor is installed at each socket of the charging strip to monitor current and voltage in real time. The high-frequency sensor synchronously samples the current and voltage at each socket. The sampling frequency must be high enough (e.g., above 1kHz) to capture instantaneous changes in current and voltage. The sensor outputs the collected current and voltage signals as analog signals. For example, suppose the charging strip has three sockets, connected to a laptop, a mobile phone charger, and a desk lamp. The high-frequency sensor samples the current and voltage at each socket at a frequency of 1kHz, generating analog signals: Socket 1: current signal It, voltage signal Vt. A high-resolution analog-to-digital converter (e.g., 12-bit or 16-bit) is selected to ensure conversion accuracy. The analog signals It and Vt are converted to digital signals using the analog-to-digital converter to obtain discrete current and voltage data. For example, after analog-to-digital conversion, the current and voltage signals of socket 1 are converted into digital signals: current data: It[n] = [0.5, 0.6, 0.55, 0.58, ...], voltage data: Vt[n] = [220, 219, 221, 220, ...].

[0018] Through frequency or time domain analysis, noise components (such as high-frequency noise and power frequency interference) in the signal are identified. Digital filtering algorithms (such as low-pass filtering and wavelet transform) are used to denoise the signal, preserving the useful signal. For example, the current signal It[n] at socket 1 is low-pass filtered to remove high-frequency noise, resulting in denoised current data: It'[n] = [0.51, 0.59, 0.56, 0.57, ...]. Based on the denoised current and voltage data, the instantaneous power of each socket is calculated and integrated to obtain the power consumption sequence for each socket.

[0019] S102. Based on the power usage sequence of each socket, a weighted average algorithm is used to calculate the load ratio of each socket relative to the total load. If the load ratio is greater than a first preset threshold, the socket corresponding to the load ratio is determined to have an abnormal distribution, and abnormal distribution characteristics of the socket corresponding to the abnormal distribution are extracted; In this embodiment, the weight of each socket is determined based on the historical usage frequency and rated power of each socket; the actual power consumption of the socket is calculated based on the weight and the total power consumption, and the load ratio of the socket relative to the total load is obtained based on the actual power consumption; the load ratio of each socket is compared with a first preset threshold value, and if the load ratio is greater than the threshold value, it is determined that the socket has abnormal distribution; for the socket with abnormal distribution, the abnormal distribution characteristics of the socket are extracted.

[0020] In this embodiment, power consumption sequence data is obtained from each socket, and noise is removed through preprocessing to obtain a standardized power consumption sequence. A weighted average algorithm is used to calculate the proportion of each socket in the total load to obtain a load proportion distribution. By comparing the load proportion with a preset threshold range, it is determined whether there is an abnormal distribution, and an abnormal marking result is obtained. If there is an abnormal distribution, the power consumption sequence characteristics of the abnormal socket are extracted to obtain an abnormal distribution characteristic. Based on the abnormal distribution characteristics, the balance of power consumption distribution among multiple sockets is analyzed to obtain a balance assessment result. A clustering algorithm is used to group the balance assessment results to obtain a power distribution pattern. By analyzing the power distribution pattern, an optimized adjustment plan for the load proportion of each socket is determined to obtain an adjusted load proportion distribution.

[0021] For example, preprocessing is a key step in acquiring electricity usage data from each outlet. Preprocessing can remove spikes and noise from the data using a median filter. Suppose the electricity usage data from a particular outlet contains a sudden outlier, such as a sudden jump to 100 amperes, while the normal value is around 10 amperes. Median filtering replaces this outlier with the median of adjacent data points, generating a smoothed series. This method preserves data trends and avoids noise interference.

[0022] In one possible implementation, standardized power usage series can be achieved through normalization. For example, the power usage of each outlet can be mapped to a range of 0 to 1. For example, if the maximum power usage of a particular outlet sequence is 500 watts and the minimum is 50 watts, normalization can map any value, such as 300 watts, to 0.56. This normalization facilitates subsequent comparisons across outlets and ensures data consistency. The core step is calculating the load ratio using a weighted average algorithm.

[0023] Specifically, a weight can be assigned to each socket, such as according to the socket priority or device type. Preferably, assuming that the system has 3 sockets with weights of 0.5, 0.3, and 0.2 respectively, the total power consumption is 1000 watts, and the actual power consumption of each socket is 400 watts, 300 watts, and 200 watts, then the proportions are calculated as 40%, 30%, and 20%. This method can reflect the actual contribution of the socket to the total load. Comparing the load ratio with the preset threshold range can identify abnormal distribution. In one embodiment, the normal ratio range is set to 30% to 50%. If a socket is only 20%, it is marked as abnormal. This judgment can quickly locate sockets with uneven power distribution.

[0024] A certain outlet has an excessively high proportion due to being connected to a high-power device, requiring further analysis. Statistical analysis can be used to extract the characteristics of the power usage series for this abnormal outlet. As you can understand, power usage series characteristics include mean and variance. Assume the mean of the abnormal outlet series is 600 watts, and the variance is large, indicating significant fluctuations in power usage. This helps determine whether the anomaly is caused by equipment failure or improper use. Analyzing the balance of power usage across multiple outlets requires a comprehensive assessment. For example, the variance of the proportions of each outlet can be calculated to assess balance. A smaller variance indicates a more balanced distribution. Assume the variance of the proportions of the three outlets is 0.01, indicating a relatively balanced distribution. This analysis can provide a basis for optimization. The use of a clustering algorithm to group the balance assessment results can identify power usage patterns. In one embodiment, the K-means algorithm is used to divide the sockets into two categories: high-load and low-load. Assuming that the proportion of high-load sockets exceeds 40% and the proportion of low-load sockets is less than 20% at a certain moment, the power usage strategy can be adjusted accordingly. This grouping can clearly present the distribution pattern of power consumption. The ultimate goal is to determine the optimization adjustment plan by analyzing the power distribution pattern. Specifically, this can be achieved by reducing the proportion of high-load sockets and increasing the utilization rate of low-load sockets. For example, after adjustment, the proportion of high-load sockets is reduced to 35%, and the low-load sockets are increased to 25%, so that the total load distribution is more even. This solution can improve system stability.

[0025] S103. Extracting the time series change rate from the abnormal distribution features to obtain a change rate sequence; obtaining a frequency component set based on the change rate sequence; separating high-frequency components from the frequency component set; calculating the oscillation amplitude of each high-frequency component to obtain an oscillation amplitude set; and, for each of the oscillation amplitude sets, marking an oscillation signal as an abnormal oscillation signal if the oscillation amplitude exceeds a preset threshold; and generating an abnormal oscillation feature set based on the abnormal oscillation signal. In this embodiment, the time series data of the power consumption of the abnormal socket is extracted from the abnormal distribution characteristics, and the rate of change between adjacent time points is calculated to obtain a rate of change sequence; the rate of change sequence is converted from the time domain to the frequency domain using a fast Fourier transform to obtain a frequency component set; components with a frequency greater than a preset threshold are screened out from the frequency component set to obtain a high-frequency component set; for each component in the high-frequency component set, its oscillation amplitude is calculated to obtain an oscillation amplitude set; the oscillation amplitude set is traversed, and if an oscillation amplitude is greater than a preset oscillation threshold, it is marked as an abnormal oscillation signal, and all the abnormal oscillation signals are aggregated to generate an abnormal oscillation feature set.

[0026] In this embodiment, the time series slope is extracted from the load ratio distribution, and the slope value at each time point is calculated to obtain a slope sequence. The slope sequence is processed using a fast Fourier transform to decompose the frequency components and obtain a frequency component set. The high-frequency components are separated from the frequency component set, and the oscillation amplitude of each high-frequency component is calculated to obtain an oscillation amplitude set. For the oscillation amplitude set, if an oscillation amplitude exceeds a preset threshold, it is marked as an abnormal oscillation signal to obtain an abnormal oscillation signal set. Extracting the time series slope from the load ratio distribution is a key step in analyzing the dynamic changes in electricity consumption. The slope reflects the rate of change of the load ratio over time and can reveal the speed of the electricity consumption trend.

[0027] For example, in a multi-outlet power system, if the load ratio of a particular outlet increases from 20% to 30% within 10 minutes, the calculated slope is 1% per minute. This slope extraction can intuitively demonstrate the dynamic characteristics of power load and facilitate subsequent analysis.

[0028] In one possible implementation, calculating the slope values at each time point to form a slope sequence requires differencing the time series. For example, suppose a socket records the load ratio once a minute. For five consecutive minutes, the data is 20%, 22%, 25%, 24%, and 26%. Differencing adjacent time points yields a slope sequence of 2%, 3%, -1%, and 2%. This method is simple and intuitive, suitable for quickly generating slope sequences and providing basic data for frequency analysis.

[0029] Preferably, the slope sequence can be processed by fast Fourier transform (FFT) to convert it from the time domain to the frequency domain and decompose the frequency components. The core of FFT is to decompose complex time changes into oscillation components of different frequencies.

[0030] For example, a slope series might contain low-frequency components reflecting steady trends, while high-frequency components correspond to rapid fluctuations. Suppose, after processing, a set of frequency components is obtained, where the low-frequency components reflect daily cyclical changes in electricity consumption, while the high-frequency components capture minute-level mutations. This decomposition method clearly reveals both the periodicity and suddenness of load changes.

[0031] As you can see, separating high-frequency components from a set of frequency components is a key step in locating rapid changes. High-frequency components are often associated with short-term anomalies in the power system. For example, a sudden and dramatic fluctuation in the load percentage of a particular outlet due to device startup would manifest as a high-frequency component.

[0032] In one embodiment, a set of high-frequency components is isolated by setting a frequency threshold, such as an oscillation above 5 times per minute. This separation can focus on rapid changes that may affect system stability. For example, calculating the oscillation amplitude of each high-frequency component can quantify the intensity of the rapid change. The oscillation amplitude reflects the degree of fluctuation of the high-frequency component in the load ratio. Assuming that the amplitude of a high-frequency component is 5%, it means that the load ratio may fluctuate by 5 percentage points in a short period of time. This quantitative method can intuitively measure the impact of fluctuations on the system.

[0033] In one possible implementation, the amplitudes of multiple high-frequency components are analyzed to form a set of oscillation amplitudes, such as 3%, 5%, and 2%, to facilitate subsequent anomaly detection.

[0034] It's important to note that determining abnormal oscillation signals requires a preset threshold for the oscillation amplitude set. For example, if the amplitude threshold is set to 4%, then components with an amplitude of 5% will be marked as abnormal oscillation signals. This marking method can quickly identify fluctuations that may be caused by equipment anomalies or improper use.

[0035] For example, if the amplitude of a certain socket exceeds the limit due to frequent starting and stopping of the motor, the device status can be further checked after marking.

[0036] In one embodiment, the formation of an abnormal oscillation signal set relies on the aggregation of all components exceeding a threshold. For example, if three high-frequency components with amplitudes exceeding the thresholds of 5%, 4.5%, and 6% are detected within a certain period, all of which are classified as abnormal oscillation signals. This aggregation provides a precise target for subsequent fault diagnosis, helping to quickly locate the problematic socket.

[0037] Specifically, the method described above forms a complete chain of logic progression from slope extraction to anomaly flagging. The slope sequence captures dynamic changes, the Fast Fourier Transform decomposes the frequency characteristics, the high-frequency components are separated to focus on rapid fluctuations, the oscillation amplitude quantifies the degree of anomaly, and finally, the anomaly signal is flagged. This layered, in-depth analysis effectively improves the accuracy of power distribution anomaly detection and provides a reliable basis for system optimization.

[0038] S104. Classify the abnormal oscillation feature set to obtain a classification result; based on the classification result and historical statistical data, calculate a probability value of the current oscillation being abnormal; if the probability value is higher than a second preset threshold, determine that the current oscillation is in an abnormal oscillation state; In an embodiment of the present application, key features are extracted from the abnormal oscillation feature set, and classification is performed based on the key features using a machine learning method. Combined with historical data statistics, the abnormal probability corresponding to each classification result is calculated; weights are assigned according to the importance of the features, and the weighted abnormal probability is calculated; the weighted abnormal probability is compared with a second preset threshold value, and if the probability value is higher than the second preset threshold value, it is determined that the current oscillation is an abnormal oscillation state.

[0039] In the embodiments of the present application, key features are extracted from the abnormal oscillation feature set. These features include oscillation amplitude, frequency components, and change trends. For example, assuming the abnormal oscillation feature set is as follows: Feature 1: Oscillation amplitude = [0.5, 0.6, 0.7, 0.8]; Feature 2: Frequency components = [50Hz, 60Hz, 70Hz, 80Hz]; Feature 3: Change trend = [increase, decrease, increase, decrease], the key features extracted are: oscillation amplitude and frequency components.

[0040] Based on the extracted key features, machine learning methods (such as support vector machines and random forests) are used to classify abnormal oscillation features. For example, using a random forest classifier to classify features, the classification results are: Classification result 1: normal oscillation, -Classification result 2: abnormal oscillation; Based on historical data statistics, calculate the abnormal probability corresponding to each classification result. For example: -Historical data statistics show that the abnormal probability of classification 1 (normal oscillation) is 0.1, and the abnormal probability of classification 2 (abnormal oscillation) is 0.9; Assign weights based on the importance of the features and calculate the weighted anomaly probability. For example, assume that the weight of feature 1 (oscillation amplitude) is 0.6 and the weight of feature 2 (frequency component) is 0.4. Calculate the weighted anomaly probability: {weighted anomaly probability} = 0.6*0.9 + 0.4*0.9 = 0.54 + 0.36 = 0.90; The weighted abnormal probability is compared with a second preset threshold. If the probability value is higher than the threshold, the current oscillation is determined to be abnormal. For example, assuming the second preset threshold is 0.8, if the comparison result is 0.90 > 0.8, the current oscillation is determined to be abnormal.

[0041] S105. Use a decision tree algorithm to obtain the correlation strength value between the abnormal oscillation state and the load ratio, perform incremental detection analysis on the correlation strength value, and obtain an incremental change sequence. If there is a value in the incremental change sequence that exceeds a third preset threshold, an overload warning signal is generated.

[0042] In an embodiment of the present application, the historical load ratio and abnormal oscillation state of the socket are combined into a data set to train a decision tree model, and the correlation strength value between the current abnormal oscillation state and load ratio of the socket is obtained based on the decision tree model; for the correlation strength value, the changing trend of the strength value is calculated to obtain an incremental change sequence; if there is a value in the incremental change sequence that exceeds a third preset threshold, an overload warning signal is generated.

[0043] In this embodiment, the historical load ratios and corresponding abnormal oscillation states of the sockets are combined into a dataset as input to the decision tree model. For example, assuming the historical data is as follows: load ratios: [0.7, 0.75, 0.8, 0.85, 0.9], and abnormal oscillation states: [0, 0, 1, 1, 1] (0 indicates normal, 1 indicates abnormal), the two are combined to form a dataset; A decision tree model was trained using the dataset to obtain the correlation strength between the load ratio and abnormal oscillation conditions. The trained decision tree model was able to predict abnormal oscillation conditions based on the load ratio. Assuming the correlation strength values output by the model are [0.2, 0.3, 0.6, 0.7, 0.8], incremental detection analysis was performed on the correlation strength values, calculating the change between adjacent strength values to obtain an incremental change sequence. Based on the incremental change formula: {incremental change} = {current strength value} - {previous strength value}, the calculated incremental change sequence is: [0.1, 0.3, 0.1, 0.1]; Check whether there is a value in the incremental change sequence that exceeds the third preset threshold. If so, generate an overload warning signal. Assume that the third preset threshold is 0.2, and the incremental change sequence is: [0.1, 0.3, 0.1, 0.1] Comparison result: 0.3 in the incremental change sequence exceeds the threshold, generating an overload warning signal.

[0044] Through the above steps, the correlation strength value between the abnormal oscillation state and the load ratio can be obtained based on the decision tree algorithm, incremental detection analysis can be performed, and an overload warning signal can be generated according to the preset threshold to effectively prevent overload risks.

[0045] The embodiment of the present invention discloses an intelligent overload warning method for a charging power strip. The method collects current and voltage signals in real time through high-frequency sensors and analog-to-digital conversion technology, and combines denoising processing and power consumption series analysis to achieve accurate calculation of the load ratio of each socket and identification of abnormal distribution. The time series change rate and frequency component are extracted from the abnormal distribution characteristics, the high-frequency component is separated and the oscillation amplitude is calculated, the abnormal oscillation signal is effectively identified, and an abnormal oscillation feature set is generated. A decision tree algorithm is used to analyze the correlation strength value between the abnormal oscillation state and the load ratio, and the overload risk is dynamically judged through incremental detection analysis to generate timely and accurate overload warning signals. The abnormal probability is calculated based on the classification results and historical statistical data, and the abnormal oscillation state is judged in combination with the preset threshold to achieve intelligent overload warning and risk prevention. Through analog-to-digital conversion, denoising processing, feature extraction and machine learning algorithms, efficient processing and analysis of complex data are achieved, thereby improving the overall performance of the system.

[0046] An embodiment of the present invention further provides an intelligent overload warning system based on a charging socket, the system comprising: Socket power consumption sequence acquisition module uses a high-frequency sensor to synchronously sample the current and voltage of each socket to obtain an analog signal, uses analog-to-digital conversion technology to convert the analog signal into a digital signal, performs denoising on the digital signal, and obtains the power consumption sequence of each socket based on the denoised data; abnormal distribution feature extraction module uses a weighted average algorithm to calculate the load ratio of each socket relative to the total load based on the power consumption sequence of each socket. If the load ratio is greater than a first preset threshold, it is determined that the socket corresponding to the load ratio has an abnormal distribution, and the abnormal distribution characteristics of the socket corresponding to the abnormal distribution are extracted; abnormal oscillation feature set generation module extracts the time series change rate from the abnormal distribution feature to obtain a change rate sequence, obtains a frequency component set based on the change rate sequence, separates the high-frequency component from the frequency component set, calculates the oscillation amplitude of each high-frequency component, and obtains an oscillation amplitude set. For the oscillation amplitude set, if a certain oscillation amplitude exceeds the preset threshold, it is marked as an abnormal oscillation signal, and an abnormal oscillation feature set is generated based on the abnormal oscillation signal; The abnormal oscillation state determination module classifies the abnormal oscillation feature set to obtain a classification result; based on the classification result and historical statistical data, calculates the probability value of the current oscillation being abnormal; if the probability value is higher than a second preset threshold, determines that the current oscillation is an abnormal oscillation state; the overload warning signal generation module adopts a decision tree algorithm to obtain the correlation strength value between the abnormal oscillation state and the load ratio, performs incremental detection analysis on the correlation strength value, and obtains an incremental change sequence; if there is a value in the incremental change sequence that exceeds a third preset threshold, an overload warning signal is generated.

[0047] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to specific embodiments. Obviously, many modifications and variations are possible based on the contents of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. An intelligent overload warning method for a charging socket, characterized in that: The method comprises the following steps: S101. Synchronously sampling the current and voltage at each outlet using a high-frequency sensor to obtain an analog signal, converting the analog signal into a digital signal using analog-to-digital conversion technology, denoising the digital signal, and obtaining a power consumption sequence for each outlet based on the denoised data. S102. Based on the power usage sequence of each socket, calculate the load ratio of each socket relative to the total load using a weighted average algorithm. If the load ratio is greater than a first preset threshold, determine that the socket corresponding to the load ratio has an abnormal distribution, and extract the abnormal distribution characteristics of the socket corresponding to the abnormal distribution; S103. Extracting the time series change rate from the abnormal distribution features to obtain a change rate sequence; obtaining a frequency component set based on the change rate sequence; separating high-frequency components from the frequency component set; calculating the oscillation amplitude of each high-frequency component to obtain an oscillation amplitude set; and, for each of the oscillation amplitude sets, marking an oscillation signal as an abnormal oscillation signal if the oscillation amplitude exceeds a preset threshold; and generating an abnormal oscillation feature set based on the abnormal oscillation signal. S104. Classify the abnormal oscillation feature set to obtain a classification result; based on the classification result and historical statistical data, calculate a probability value of the current oscillation being abnormal; if the probability value is higher than a second preset threshold, determine that the current oscillation is in an abnormal oscillation state; S105. Use a decision tree algorithm to obtain the correlation strength value between the abnormal oscillation state and the load ratio, perform incremental detection analysis on the correlation strength value, and obtain an incremental change sequence. If there is a value in the incremental change sequence that exceeds a third preset threshold, an overload warning signal is generated.

2. The method according to claim 1, characterized in that The S102 includes: Determine the weight of each socket according to the historical usage frequency and rated power of each socket; Calculating the actual power consumption of the socket according to the weight and the total power consumption, and obtaining the load ratio of the socket relative to the total load based on the actual power consumption; Comparing the load ratio of each socket with a first preset threshold, and if the load ratio is greater than the threshold, determining that the socket has abnormal allocation; For the sockets with abnormal distribution, abnormal distribution features of the sockets are extracted.

3. The method according to claim 1, characterized in that The S103 includes: Extract the time series data of power consumption of abnormal sockets from the abnormal distribution characteristics, calculate the change rate between adjacent time points, and obtain the change rate series; Convert the rate of change sequence from the time domain to the frequency domain using a fast Fourier transform to obtain a set of frequency components; Filtering out components with frequencies greater than a preset threshold from the frequency component set to obtain a high-frequency component set; For each component in the high-frequency component set, calculating its oscillation amplitude to obtain an oscillation amplitude set; The oscillation amplitude set is traversed, and if an oscillation amplitude is greater than a preset oscillation threshold, it is marked as an abnormal oscillation signal, and all the abnormal oscillation signals are aggregated to generate an abnormal oscillation feature set.

4. The method according to claim 3, characterized in that The S104 includes: Extract key features from the abnormal oscillation feature set, classify the data using a machine learning method based on the key features, and calculate the abnormal probability corresponding to each classification result in combination with historical data statistics; Assign weights based on the importance of features and calculate weighted anomaly probability; The weighted abnormal probability is compared with a second preset threshold value, and if the probability value is higher than the second preset threshold value, the current oscillation is determined to be an abnormal oscillation state.

5. The method according to claim 1, wherein The S105 includes: A data set is formed by combining the historical load ratio and abnormal oscillation state of the socket to train a decision tree model, and a correlation strength value between the abnormal oscillation state and the load ratio of the current socket is obtained based on the decision tree model; For the correlation strength value, calculating a change trend of the strength value to obtain an incremental change sequence; If there is a value in the incremental change sequence that exceeds a third preset threshold, an overload warning signal is generated.

6. An intelligent overload warning system for a charging socket, characterized in that: The system comprises: The module acquires the power consumption sequence of each socket by synchronously sampling the current and voltage of each socket through a high-frequency sensor to obtain an analog signal, converts the analog signal into a digital signal using analog-to-digital conversion technology, denoises the digital signal, and obtains the power consumption sequence of each socket based on the denoised data; an abnormal distribution feature extraction module, which calculates the load ratio of each socket relative to the total load using a weighted average algorithm based on the power consumption sequence of each socket, and if the load ratio is greater than a first preset threshold, determines that the socket corresponding to the load ratio has abnormal distribution, and extracts abnormal distribution features of the socket corresponding to the abnormal distribution; an abnormal oscillation feature set generation module, which extracts the time series change rate from the abnormal distribution feature to obtain a change rate sequence, obtains a frequency component set based on the change rate sequence, separates high-frequency components from the frequency component set, calculates the oscillation amplitude of each high-frequency component to obtain an oscillation amplitude set, and for the oscillation amplitude set, if an oscillation amplitude exceeds a preset threshold, marks it as an abnormal oscillation signal, and generates an abnormal oscillation feature set based on the abnormal oscillation signal; an abnormal oscillation state determination module, which classifies the abnormal oscillation feature set to obtain a classification result; calculates a probability value of the current oscillation being abnormal based on the classification result and historical statistical data; and determines that the current oscillation is in an abnormal oscillation state if the probability value is higher than a second preset threshold; The overload warning signal generation module adopts a decision tree algorithm to obtain the correlation strength value between the abnormal oscillation state and the load ratio, performs incremental detection analysis on the correlation strength value, and obtains an incremental change sequence. If there is a value in the incremental change sequence that exceeds a third preset threshold, an overload warning signal is generated.

7. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the intelligent overload warning method for the charging socket as described in any one of claims 1 to 5 is implemented.

Images