A centralized carport charging safety management and control method, device, equipment and medium

Abstract

The present application belongs to the technical field of electric bicycle charging safety monitoring, and particularly relates to a centralized carport charging safety management method, device, equipment and medium. The method comprises: acquiring electrical parameter data of each charging circuit, preprocessing to obtain a multi-dimensional feature vector of each charging circuit, the multi-dimensional feature vector at least including active power characteristics, power factor characteristics, current harmonic characteristics, temperature rise characteristics and residual current characteristics; identifying the charging state of the current charging circuit based on the multi-dimensional feature vector; calculating the comprehensive risk value of the current charging circuit based on the preset nonlinear fusion model according to the charging state of the current charging circuit and the multi-dimensional feature vector; comparing the comprehensive risk value with the preset multiple risk level thresholds, and executing the safety disposal action corresponding to the risk level according to the comparison result. The present application realizes accurate quantification and hierarchical management and control of charging risks through multi-feature fusion and nonlinear product model.

Description

Technical Field

[0001] This invention belongs to the field of electric bicycle charging safety monitoring technology, specifically relating to a centralized bicycle shed charging safety management method, device, equipment, and medium. Background Technology

[0002] Centralized carport charging scenarios present challenges such as high electrical fire risks, low efficiency of manual inspections, and susceptibility to environmental interference with traditional monitoring methods. Existing technologies often rely on single electrical quantity thresholds or video surveillance, which struggle to accurately identify charging behavior stages, resulting in high false alarm and false negative rates. Summary of the Invention

[0003] The purpose of this invention is to provide a centralized carport charging safety management method, device, equipment and medium, so as to at least solve or improve one of the problems existing in the prior art.

[0004] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for safe management and control of centralized carport charging, comprising the following steps: Obtain electrical parameter data for each charging circuit. The electrical parameter data should include at least the circuit voltage, circuit current, residual current, and temperature. The electrical parameter data is preprocessed to obtain a multi-dimensional feature vector for each charging circuit. The multi-dimensional feature vector includes at least active power characteristics, power factor characteristics, current harmonic characteristics, temperature rise characteristics, and residual current characteristics. Identify the charging state of the current charging circuit based on multi-dimensional feature vectors; Based on the current charging state of the charging circuit and the multi-dimensional feature vector, the comprehensive risk value of the current charging circuit is calculated based on a preset nonlinear fusion model. The overall risk value is compared with multiple preset risk level thresholds, and safety measures corresponding to the risk level are executed based on the comparison results.

[0005] The aforementioned solution acquires multiple electrical parameters, including loop voltage, loop current, residual current, and temperature. After data preprocessing, it extracts active power characteristics, power factor characteristics, current harmonic characteristics, temperature rise characteristics, and residual current characteristics. Based on this, it identifies the charging status and calculates a comprehensive risk value. Finally, it implements tiered safety measures according to the risk level. This solution addresses the problems of existing technologies that rely on single electrical quantity thresholds or video monitoring, have difficulty accurately identifying charging behavior stages, and suffer from high false alarm and false negative rates. It achieves multi-dimensional, multi-stage, and refined risk perception and tiered control of the charging circuit, effectively improving electrical fire prevention capabilities in carport charging scenarios.

[0006] Furthermore, the current charging state of the charging circuit is identified based on multi-dimensional feature vectors, specifically including the following steps: A finite state machine model is established using multidimensional feature vectors as input. The state set of the finite state machine model includes standby state, power-on state, constant current charging state, constant voltage charging state, end-of-life or fully charged state, and abnormal state. Obtain the set of configurable parameters issued by the platform. The set of configurable parameters includes at least the following: power-on confirmation delay window, constant current confirmation delay window, constant voltage confirmation delay window, end-of-charge confirmation delay window, anomaly confirmation delay window, upper and lower thresholds for the proportion of each harmonic, power change rate threshold, charging end power threshold, power factor change rate threshold, power decrease rate threshold, temperature difference change rate threshold, harmonic proportion sudden change threshold, temperature rise rate threshold, and remaining current threshold. Based on multidimensional feature vectors, configurable parameter sets, and preset state transition rules, the charging state and verification mark of the current time window are determined.

[0007] The above solution establishes a finite state machine model using multi-dimensional feature vectors as input and incorporates a set of configurable parameters (including various delay windows, harmonic thresholds, power change rate thresholds, etc.) issued by the platform. It determines the charging state and verification markers based on preset state transition rules. This solution solves the problem of existing technologies being unable to accurately distinguish between fine-grained charging stages such as standby, power-on, constant current, constant voltage, termination, and anomalies. This enables dynamic state tracking and anomaly verification during the charging process, providing accurate state input for subsequent risk classification.

[0008] Furthermore, the state transition rules include at least the following: When the device is in standby mode and the charging state gate signal is continuously held for a number of time windows equal to the power-on confirmation delay window, it will transition to the power-on state. When the state is in the power-on state and the proportion of each harmonic is between the corresponding upper and lower thresholds, and the absolute value of the first difference of the active power characteristic is not greater than the power change rate threshold, the state is continuously maintained for a number of time windows of constant current confirmation delay window, and then it is transferred to the constant current charging state. When the constant current charging state is maintained for a number of time windows, and the first difference of the power factor characteristic is not less than the power factor change rate threshold, and the first difference of the active power characteristic is not greater than the negative power decrease rate threshold, the constant voltage confirmation delay window state is maintained for a number of time windows. When the state of constant voltage charging, active power characteristic not greater than the charging end power threshold, and absolute value of the second difference of temperature difference characteristic not greater than the temperature difference change rate threshold is maintained for a number of time windows for the closing confirmation delay window, the state is transferred to the closing or full charge state. When the maximum value of the first-order difference of the harmonic proportion characteristic is not less than the harmonic proportion mutation threshold, or the temperature rise rate characteristic is not less than the temperature rise rate threshold, or the residual current characteristic is not less than the residual current threshold, and the state is continuously maintained for the number of time windows of the abnormal confirmation delay window, the state is transferred to the abnormal state and a verification mark is set.

[0009] The above scheme defines specific state transition rules, binding the switching conditions between standby, power-on, constant current, constant voltage, end-of-charge / fully charged, and abnormal states to threshold values ​​and delay windows of characteristic parameters (harmonic ratio, active power, power factor, temperature difference, temperature rise rate, and residual current). This scheme solves the problems of charging state transitions being susceptible to transient interference and ambiguous boundary conditions, thus achieving charging state transition control with strong anti-interference capabilities, clear boundaries, and rigorous logic.

[0010] Furthermore, the overall risk value The calculation method is as follows:

[0011] in, Configurable weighting coefficients for state factors; This is a state factor used to quantify the contribution of the current charging state to the risk. The current charging state of the charging circuit ρ; Configurable weight coefficients for the i-th type of feature; For the nonlinear mapping function of the i-th type of feature, respectively, it corresponds to the quantification of the degree of anomaly of active power feature, power factor feature, current harmonic feature, temperature rise feature and residual current feature; For the charging circuit ρ within the time window The multidimensional feature vector.

[0012] The above scheme constructs a composite risk index calculation model, fusing the state factor with five characteristic factors in a product form. Each characteristic factor quantifies the degree of anomaly of a specific physical characteristic, while the state factor reflects the basic risk of the current charging state. This scheme solves the problem of misjudgment or omission of risk due to independent judgment of a single feature, thereby achieving comprehensive risk quantification through multi-feature synergy and nonlinear superposition, improving the accuracy and sensitivity of risk assessment.

[0013] Furthermore, the overall risk value is compared with multiple preset risk level thresholds, and safety procedures corresponding to the risk level are executed based on the comparison results, including: Obtain the hierarchical parameter set issued by the platform, which includes at least: a first risk threshold, a second risk threshold, a third risk threshold, and a first-level alarm delay window, a second-level alarm delay window, and a third-level alarm delay window; wherein the first risk threshold is less than the second risk threshold, and the second risk threshold is less than the third risk threshold; Three levels of risk conditions are defined; among them, the severe risk condition is: the comprehensive risk value is greater than the third risk threshold, or the charging state is abnormal; the medium risk condition is: the comprehensive risk value is greater than the second risk threshold, or the review mark is true; the warning risk condition is: the current charging state is and the comprehensive risk value is greater than the first risk threshold. The application of a delay-hold operator determines the three levels of risk conditions. Specifically, if a severe risk condition is maintained for a number of consecutive time windows corresponding to the third-level alarm delay window, then the third-level alarm level is output; if a moderate risk condition is maintained for a number of consecutive time windows corresponding to the second-level alarm delay window, then the second-level alarm level is output; if a warning risk condition is maintained for a number of consecutive time windows corresponding to the first-level alarm delay window, then the first-level alarm level is output; otherwise, no alarm level is output. The corresponding safety procedures are executed according to the alarm level. Specifically, when the alarm level is Level 1, a prompting action is executed, including at least an audible and visual alarm; when the alarm level is Level 2, a control action is executed, including at least stopping charging or limiting the charging current; and when the alarm level is Level 3, a power-off action is executed, including at least disconnecting the power supply relay or contactor.

[0014] The above solution defines three levels of risk conditions by setting three risk thresholds and corresponding delay windows, combining the comprehensive risk value with charging status and verification flags. It then applies a delay-holding operator to determine the alarm level and finally executes a tiered safety response from alert to control to power-off based on the alarm level. This solution solves the problems of ambiguous risk classification, fluctuating alarms, and limited response actions, thus achieving clear, fluctuating, and progressively enhanced safety management, avoiding unnecessary power-off interference while ensuring safety.

[0015] Furthermore, it also includes: The comprehensive risk value, corresponding risk level, and corresponding multi-dimensional feature vector of the current charging circuit are encapsulated into a unified event frame and uploaded to the remote management platform. The remote management platform extracts multidimensional feature vectors and corresponding risk levels from multiple received event frames to construct a labeled sample set; Using the sample set as input, the configurable parameters in the nonlinear fusion model are iteratively optimized by minimizing the cross-entropy loss function with regularization. The optimized configurable parameters are sent to the front end to update subsequent risk calculations.

[0016] The above solution encapsulates the comprehensive risk value, risk level, and multi-dimensional feature vectors into a unified event frame and uploads it to a remote management platform. The platform then constructs a labeled sample set, iteratively optimizes the model parameters, and distributes the data to the front end. This solution addresses the problem of fixed model parameters that cannot adapt to different charging circuit characteristics or changing operating conditions, thereby achieving adaptive optimization and continuous iteration of the model and improving the generalization ability and accuracy of the risk identification model in different scenarios.

[0017] Furthermore, the configurable parameters in the nonlinear fusion model are iteratively optimized; the optimization problem is:

[0018] Where θ represents all parameters to be optimized; σ( () is the Sigmoid function; Let be the overall risk value of the i-th sample; The corresponding risk level threshold; This represents the true risk label for the i-th sample. The total number of samples; is the regularization coefficient.

[0019] The above scheme constructs a cross-entropy loss function with a regularization term, using a historical sample set as input, to iteratively optimize the configurable parameters in the nonlinear fusion model. This scheme solves the problems of strong subjectivity and difficulty in achieving global optimum in manual parameter tuning, thereby realizing data-driven automated parameter optimization, reducing the cost of manual intervention, and improving the accuracy and robustness of the model.

[0020] In a second aspect, the present invention provides a centralized carport charging safety management and control device, comprising: The data acquisition module is used to acquire electrical parameter data for each charging circuit. The electrical parameter data includes at least circuit voltage, circuit current, residual current, and temperature. The data processing module is used to preprocess electrical parameter data to obtain a multi-dimensional feature vector for each charging circuit. The multi-dimensional feature vector includes at least active power characteristics, power factor characteristics, current harmonic characteristics, temperature rise characteristics, and residual current characteristics. The status recognition module is used to identify the current charging state of the charging circuit based on multi-dimensional feature vectors. The risk identification module is used to calculate the comprehensive risk value of the current charging circuit based on the charging status of the current charging circuit and multi-dimensional feature vectors, using a preset nonlinear fusion model. The instruction generation module is used to compare the comprehensive risk value with multiple preset risk level thresholds, and execute safety handling actions corresponding to the risk level based on the comparison results.

[0021] In a third aspect, the present invention provides an electronic device including a processor and a memory, the processor being configured to execute a computer program stored in the memory to implement the centralized carport charging safety management method as described above.

[0022] In a fourth aspect, the present invention provides a computer-readable storage medium storing at least one instruction that, when executed by a processor, implements the centralized vehicle shed charging safety management method as described above. Attached Figure Description

[0023] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a flowchart illustrating a centralized carport charging safety management method according to an embodiment of the present invention; Figure 2 This is a structural block diagram of a centralized carport charging safety management and control device according to an embodiment of the present invention; Figure 3 This is a structural block diagram of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0024] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0025] The following detailed description is exemplary and intended to provide further detailed explanation of the invention. Unless otherwise specified, all technical terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. The terminology used in this invention is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention.

[0026] Example 1 like Figure 1 As shown, a centralized vehicle shed charging safety management method includes the following steps: S1. Obtain electrical parameter data for each charging circuit. The electrical parameter data shall include at least the circuit voltage, circuit current, residual current, and temperature.

[0027] S2. Perform data preprocessing on the electrical parameter data to obtain a multi-dimensional feature vector for each charging circuit. The multi-dimensional feature vector includes at least active power characteristics, power factor characteristics, current harmonic characteristics, temperature rise characteristics, and residual current characteristics.

[0028] S3. Identify the charging state of the current charging circuit based on multi-dimensional feature vectors; Specifically, a finite state machine model is established using multidimensional feature vectors as input. The state set of the finite state machine model includes standby state, power-on state, constant current charging state, constant voltage charging state, end-of-life or fully charged state, and abnormal state. Obtain the set of configurable parameters issued by the platform. The set of configurable parameters includes at least the following: power-on confirmation delay window, constant current confirmation delay window, constant voltage confirmation delay window, end-of-charge confirmation delay window, anomaly confirmation delay window, upper and lower thresholds for the proportion of each harmonic, power change rate threshold, charging end power threshold, power factor change rate threshold, power decrease rate threshold, temperature difference change rate threshold, harmonic proportion sudden change threshold, temperature rise rate threshold, and remaining current threshold. Based on multidimensional feature vectors, configurable parameter sets, and preset state transition rules, the charging state and verification mark of the current time window are determined.

[0029] The state transition rules include at least the following: When the device is in standby mode and the charging state gate signal is continuously held for a number of time windows equal to the power-on confirmation delay window, it will transition to the power-on state. When the state is in the power-on state and the proportion of each harmonic is between the corresponding upper and lower thresholds, and the absolute value of the first difference of the active power characteristic is not greater than the power change rate threshold, the state is continuously maintained for a number of time windows of constant current confirmation delay window, and then it is transferred to the constant current charging state. When the constant current charging state is maintained for a number of time windows, and the first difference of the power factor characteristic is not less than the power factor change rate threshold, and the first difference of the active power characteristic is not greater than the negative power decrease rate threshold, the constant voltage confirmation delay window state is maintained for a number of time windows. When the state of constant voltage charging, active power characteristic not greater than the charging end power threshold, and absolute value of the second difference of temperature difference characteristic not greater than the temperature difference change rate threshold is maintained for a number of time windows for the closing confirmation delay window, the state is transferred to the closing or full charge state. When the maximum value of the first-order difference of the harmonic proportion characteristic is not less than the harmonic proportion mutation threshold, or the temperature rise rate characteristic is not less than the temperature rise rate threshold, or the residual current characteristic is not less than the residual current threshold, and the state is continuously maintained for the number of time windows of the abnormal confirmation delay window, the state is transferred to the abnormal state and a verification mark is set.

[0030] S4. Based on the charging status of the current charging circuit and the multi-dimensional feature vector, calculate the comprehensive risk value of the current charging circuit based on the preset nonlinear fusion model.

[0031] S5. Compare the overall risk value with multiple preset risk level thresholds, and execute the safety handling actions corresponding to the risk level based on the comparison results.

[0032] In one embodiment, an electrical safety monitoring unit is deployed on each charging circuit. The electrical safety monitoring unit includes at least a voltage transformer, a current transformer, a residual current transformer, and a temperature probe.

[0033] In one embodiment, after acquiring the electrical parameter data of each charging circuit, the method further includes: establishing a mapping relationship between the monitoring channel and the physical port of each charging circuit. ,in, For monitoring channel aggregation, A set of physical ports; through mapping relationships The real-time collected electrical parameter data is assigned to the corresponding charging circuit ρ, forming the original discrete time-series data of each charging circuit.

[0034] Set the sampling frequency f s Power frequency f0 = 50Hz, f s Take an integer multiple of f0 and define: , , (1) in, denoted as the power frequency angular frequency; N is the number of sampling points per power frequency cycle; L is the window length covering m power frequency cycles; m is a positive integer representing the number of power frequency cycles covered by the sliding window.

[0035] For the charging circuit ρ, through the mapping relationship The original discrete time series is obtained from the channel data. : , (2) in, Indicates from physical port The collected raw electrical parameter values ​​corresponding to the monitoring channel γ at sampling time n.

[0036] Original Discrete Time Series Equivalently written as the original multi-channel sample vector : (3) in, This represents the original multi-channel sample vector of the charging circuit ρ at sampling time n; Let ρ be the loop voltage of the charging circuit at sampling time n; Let ρ be the loop current of the charging circuit at sampling time n; Let ρ be the residual current of the charging circuit at sampling time n; ~ The temperature of four different temperature measurement points (cable, terminal block, charging gun head, and environment) deployed on the charging circuit ρ.

[0037] In one embodiment, data preprocessing of electrical parameter data includes: Filtering and denoising the original multi-channel sample vector includes: filtering the loop voltage. Loop current The sequence is subjected to notch or bandpass filtering for temperature. With residual current Perform moving average or median filtering on the sequence to obtain a denoised sequence. .

[0038] Then in each time window Internally, based on the loop voltage after filtering and noise reduction Loop current Calculate the average active power and power factor According to average active power and power factor Determine active power characteristics Power factor characteristics .

[0039] (4) (5) Where Δ is the sliding step size, Δ≤L; These are the effective values ​​of voltage and current within the corresponding time window: , (6) Extracting current harmonic features, including the proportion of the third harmonic. 5th harmonic proportion characteristics 7th harmonic proportion characteristics Specifically, a Fast Fourier Transform (FFT) is performed on the current sequence within the window to obtain the frequency domain components, and the amplitudes of the r ∈ {3, 5, 7} harmonics are extracted. and its proportion : (7) , (8) in, is the Discrete Fourier Transform (DFT) result of the current sequence, and s is the discrete frequency index in the frequency domain. When s=1, it corresponds to the fundamental component, and when s=3, 5, and 7, it corresponds to the 3rd, 5th, and 7th harmonic components, respectively. The amplitude is the fundamental wave.

[0040] Extract temperature rise features, including temperature difference features. and temperature rise rate characteristics Specifically, firstly, for each temperature measurement point k, its time window is defined. The representative temperature value inside For example, take the arithmetic mean or median of the temperature sequence at the measurement point within the window; then define the maximum, minimum, and temperature difference within the window: , (9) (10) in, This represents the maximum temperature difference within the loop, used to characterize the uniformity of temperature distribution within the loop. According to... Determine temperature difference characteristics .

[0041] The rate of temperature rise (the rate of change of temperature difference) is defined as: (11) in, The time interval between adjacent time windows, for example , The sampling period; The rate of temperature rise, or the rate of change of temperature difference relative to the previous time window, is used to characterize the drasticness of temperature change. According to... Determine the characteristics of the temperature rise rate .

[0042] The residual current characteristics are determined based on the average or effective value of the residual current within the window. .

[0043] Optionally, some features can be normalized using Z-score or Min-Max, but the original values ​​are retained for the physical thresholds used in subsequent grading.

[0044] Finally, a loop-level multidimensional feature vector is obtained. : (12) And calculate First-order difference and second-order difference This is used for subsequent state transition judgments.

[0045] In one embodiment, multidimensional feature vectors Specifically, this includes: active power characteristics Power factor characteristics Characteristics of the proportion of third harmonics 5th harmonic proportion characteristics 7th harmonic proportion characteristics Residual current characteristics Temperature difference characteristics and temperature rise rate characteristics ;in, This is the time window index, and the subscript ρ is the charging circuit number.

[0046] In one embodiment, identifying the current charging state of the charging circuit based on multi-dimensional feature vectors specifically includes: Based on multidimensional feature vectors, the charging state is identified using a finite state machine model. The state set S of the finite state machine model includes standby state S0, power-on state S1, constant current charging state S2, constant voltage charging state S3, end-of-charge / fully charged state S4, and abnormal state S5. x .

[0047] For each charging loop ρ, a multidimensional feature vector is used. A finite state machine is built using the input. The transition conditions of the finite state machine are determined by a set of configurable parameters issued by the platform. Parameterize and output the current state of the charging loop ρ. and verification mark .

[0048] Introducing a comprehensive score Used for charging gating determination: (13) Among them, weight These are configurable parameters, namely the weight coefficients of each feature factor. Corresponding active power characteristic weights, Corresponding power factor characteristic weights, Corresponding to the comprehensive characteristic weights of the third / fifth / seventh harmonics, Corresponding to the comprehensive characteristic weight of temperature rise, Corresponding residual current characteristic weights; function This is the feature normalization function, which maps each feature to the interval [0,1].

[0049] Specifically, The power can be mapped to [0,1] using a sigmoid function. Normalize the power factor The maximum value of the three harmonic proportions or the weighted sum can be taken and then normalized. , Normalize the temperature difference and residual current separately. The normalization method can be selected according to the actual data distribution, such as Min-Max or Z-Score.

[0050] Charging status gate signal Defined as: (14) in, For indicator functions, A threshold for determining the charging status. This indicates that the device is charging; otherwise, it indicates that it is not charging.

[0051] To suppress transient disturbances, a delay-and-hold operator is introduced. For any Boolean condition If it remains constant at 1 for δ consecutive time windows, then it is denoted as Otherwise, it is 0.

[0052] Configurable parameter set issued by the platform : (15) in, This is a power-on confirmation delay window used to filter out transient interference while the charging state continues. Only after a window is opened is it considered to have truly entered the power-on state.

[0053] This is a constant current confirmation delay window, used to confirm entry into the constant current charging phase.

[0054] This is a constant voltage confirmation delay window, used to confirm entry into the constant voltage charging stage.

[0055] This is a confirmation delay window for the finalization process, used to confirm entry into the finalization / full-up phase.

[0056] This is an exception confirmation delay window used to filter out instantaneous exception spikes, and the exception state is only triggered after a continuous exception is confirmed.

[0057] , These are the lower and upper thresholds for the proportion of the third harmonic, used to determine whether the harmonics are within the normal range.

[0058] , These are the lower and upper thresholds for the proportion of the 5th harmonic, respectively.

[0059] , These are the lower and upper thresholds for the proportion of the 7th harmonic, respectively.

[0060] This is the power change rate threshold, used to determine whether the power has entered a steady state.

[0061] This is the power threshold for ending charging. When the power is below this value for a period of time, charging is considered to have ended.

[0062] This is the power factor change rate threshold, used to identify characteristic changes during the transition from constant current to constant voltage.

[0063] This is the power decay rate threshold, used to identify characteristics of power decay during the constant voltage phase.

[0064] This is the threshold for the rate of change of temperature difference, used to determine whether the temperature tends to stabilize.

[0065] This is the threshold for abrupt changes in the harmonic proportion. When the change in the harmonic proportion within a window exceeds this value, it is considered abnormal.

[0066] This is the temperature rise rate threshold; when the temperature rise rate exceeds this value, it is considered abnormal.

[0067] This is the residual current threshold. When the residual current exceeds this value, it is considered an abnormal leakage current.

[0068] Predefined first-order difference of power factor characteristics: .

[0069] The transition rules for the finite state machine model are defined as follows (deterministic rules): Standby → Power On: (16) in, For power-on delay hold operator, This is the index for the next time window.

[0070] Power on → Constant current: (17) in, For constant current delay sustain operator; The first-order difference represents the active power characteristic; Constant current → Constant voltage: (18) in, For constant voltage delay hold operator; and These are the first-order differences for power factor characteristics and active power characteristics, respectively.

[0071] Constant pressure → Finishing / Full charge: (19) in, To maintain the operator for the final delay, It is a second-order difference characteristic of temperature difference.

[0072] Any state → Anomaly (with verification flag): Triggered when abnormal waveforms, abnormal temperature rise, or abnormal residual current occur. And mark it as reviewed: (20) , (twenty one) in, For abnormal delay hold operators; The first-order difference represents the harmonic proportion characteristic.

[0073] The platform can also perform binary classification logistic regression on the labeled sample set, optimizing α and α. Wait, get the updated parameters and use them as A subset is distributed to the front end for continuous optimization.

[0074] In one embodiment, the nonlinear fusion model is specifically a comprehensive risk value. The calculation method is as follows: (twenty two) in, For the charging circuit ρ within the time window The multidimensional feature vector; The current charging state of the charging circuit ρ; Configurable weight coefficients for the i-th type of feature (i=1,…,5); Configurable weighting coefficients for state factors; For the nonlinear mapping function of the i-th type of feature, respectively, it corresponds to the quantification of the degree of anomaly of active power feature, power factor feature, current harmonic feature, temperature rise feature and residual current feature; This is a state factor used to quantify the contribution of the current charging state to the risk.

[0075] It should be noted that the nonlinear fusion model adopts a product-based fusion framework, where each feature factor ( This function independently reflects the risk contribution of a specific type of anomaly (active power, power factor, harmonics, temperature rise, residual current), and its factor value is always greater than or equal to 1. The product operation accurately characterizes the synergistic amplification effect when multiple anomalies occur concurrently: the risk moderately increases when a single anomaly exists, but when multiple anomalies coexist, the risk index rises sharply due to the product effect, consistent with the physical laws of hazard accumulation and multi-factor coupling in actual electrical fire risks. Furthermore, each nonlinear mapping function... All follow the mathematical properties of being monotonically increasing, bounded, or semi-bounded, with the temperature rise factor taking an exponential form. It can reflect the dual accelerating effect of temperature rise and temperature rise rate on fire risk, which is consistent with the physical mechanism of thermal runaway; power factor, power factor and harmonic factor are adopted The function possesses saturation properties, which can prevent a single feature from excessively dominating the risk value, enabling the model to have soft-limiting capabilities against extreme anomalies. State factor The product-based summation reflects the amplifying effect of the charging state on the basic risk, ensuring that the risk assessment has reasonable benchmark differences at different charging stages.

[0076] Specifically, five nonlinear mapping functions are defined as follows: Power Factor This is used to reflect the degree to which the charging power deviates from the normal range, and is expressed as: (twenty three) in, The sensitivity coefficient of the power factor. , This is the upper limit safety threshold of active power for the charging circuit ρ to operate normally.

[0077] Power factor This is used to reflect the risk of an excessively low power factor, and is expressed as: (twenty four) in, The sensitivity coefficient of the power factor. , ρ is the lower limit threshold of the power factor under normal charging conditions in the charging circuit.

[0078] Harmonic factors This is used to reflect the degree of abnormality in the proportion of the 3rd, 5th, and 7th harmonics, and is expressed as: (25) in, The weighting coefficient for the r-th harmonic is... ; Let be the sensitivity parameter for the r-th harmonic. The threshold value for the rth harmonic is given.

[0079] Temperature rise factor This is used to reflect the combined effect of temperature rise and its rate, and is expressed as: (26) in, This is the temperature difference weighting coefficient. >0 represents the temperature rise rate weighting coefficient, which is used to adjust the contribution of temperature difference and temperature rise rate to the temperature rise factor.

[0080] Leakage factor The square effect of residual current is used to reflect this effect and is expressed as: (27) in, , is the global coefficient of the leakage factor.

[0081] Specifically, state factors This is used to quantify the contribution of the current charging state to the risk, and is expressed as: (28) Among them, standby state S0 and the finishing / fully charged state S4 are considered low-risk states. Power-on state S1, constant current charging state S2, and constant voltage charging state S3 represent normal charging processes, assigned a medium base risk value of 0.3; abnormal state S... x Assign the highest risk value of 1. This mapping can be adjusted according to the actual scenario, 0 ≤ ≤ 1.

[0082] all The range of each factor is [0, +∞). Overall risk value When all features are normal and the charging state is standby or nearing completion, (i=1,…,5) and , When a certain feature is abnormal, the corresponding factor increases. Due to the product effect, multiple abnormalities will cause the exponent to rise sharply.

[0083] The platform distributes and maintains configurable parameter sets online. , in, As the risk level threshold, Maintain the number of windows for delays at all levels of alarms.

[0084] Based on the comprehensive risk value and state identification output, three types of classification conditions are defined: L3 (Severe) Conditions: (29) L2 (Medium) conditions: (30) L1 (Hint): (31) Apply the delay-and-hold operator to determine the final hierarchical output: (32) in, , , These represent the time window. Below, the Boolean flag indicates whether the charging circuit ρ meets the three risk levels of severe, moderate, and warning. Indicates no alarm output, project field Corresponding to This means that when multiple conditions are met simultaneously, the system outputs the alarm level with the highest priority. Output the final alarm level.

[0085] Three distinct types of execution outputs are defined for each charging loop ρ: This is a prompt output; a value of 1 indicates that a prompt action such as an audible or visual alarm will be executed. This is a control output. A value of 1 indicates that a control action of stopping charging or limiting the charging current is performed. The execution output, when the value is 1, indicates that the power supply relay or contactor is disconnected, thus physically cutting off the power.

[0086] Action mapping is defined as: (33) And stipulate that when When power is off, output It has the highest priority and overrides other outputs.

[0087] In one embodiment, the preset multiple risk level thresholds include a first risk threshold. Second risk threshold and the third risk threshold ,and < < ; Based on the comparison results, implement safety procedures corresponding to the risk level, specifically including: When the comprehensive risk value ≥Third risk threshold When this occurs, a Level 3 safety response should be initiated, which includes at least cutting off the power supply. When the second risk threshold ≤Comprehensive Risk Value Third risk threshold When this occurs, Level 2 safety procedures should be implemented, which include at least current limiting or stopping charging. When the first risk threshold ≤Comprehensive Risk Value Second risk threshold When the current charging circuit is in the charging state, the first level of safety action is executed, which includes at least an audible and visual prompt.

[0088] Based on the aforementioned hierarchical logic, the above actions correspond to The output, and based on the delay hold window Make a judgment to ensure the reliability of the action.

[0089] In one embodiment, the method further includes: The comprehensive risk value, corresponding risk level, and corresponding multidimensional feature vector of the current charging circuit are encapsulated into a unified event frame and uploaded to the remote management platform. The remote management platform extracts the multidimensional feature vector and corresponding risk level from the received event frames to construct a labeled sample set, where each sample contains a multidimensional feature vector and its true risk label. Using the sample set as input, the configurable parameters in the nonlinear fusion model are iteratively optimized by minimizing the cross-entropy loss function with a regularization term. The configurable parameters include configurable weight coefficients. and nonlinear mapping functions The internal parameters in the system; the optimized configurable parameters are distributed; the optimized configurable parameters are received and the updates are applied to optimize subsequent risk calculations.

[0090] Specifically, define loop-level event frames. for: (34) Where ID is the device identifier and ρ is the loop number. For mapping Version number, For hierarchical fields, corresponding ; For event timestamps, The version of the parameter set that was in effect at that time. For the graded parameter set that was in effect at the time, the load data Defined as: (35) in, For window index set, This is a validation field. If the platform does not perform a validation within the timeout T... ack If an internal confirmation is received, the same frame will be retransmitted, with a maximum of M retransmissions. max If a retransmission fails, it is stored in the local buffer queue Q and retransmitted after the link is restored.

[0091] The platform optimizes the parameters and risk level thresholds in the comprehensive risk value based on historical event data. The optimization problem can be transformed into: (36) Where θ represents all parameters to be optimized, such as , , wait; This represents the true risk label for the i-th sample. Let be the comprehensive risk value of the i-th sample (from charging circuit ρ); The regularization coefficient is used. σ( is the total number of samples; () is the Sigmoid function. The threshold values ​​are for the corresponding levels. The updated parameter set is obtained by solving using gradient descent or a quasi-Newton method. .

[0092] It should be noted that the optimization problem described above uses a binary classification cross-entropy loss function, which calculates the difference between the continuous risk index and the level threshold. The risk model parameters are mapped to probability values ​​using the Sigmoid function, representing the probability that a sample's risk exceeds a certain threshold. This transforms risk model parameter optimization into a probabilistic classification task, yielding a clear gradient signal. An L2 regularization term is then introduced. This approach prevents overfitting and ensures the model's generalization ability even with a limited sample size. The optimization variables encompass all configurable parameters in the model (including weight coefficients, sensitivity coefficients, etc.), while the risk level threshold is set independently as an engineering parameter. This decoupling allows system deployment and maintenance personnel to flexibly adjust the parameters based on on-site tolerance. Through data-driven end-to-end optimization, it overcomes the shortcomings of traditional methods that rely on human experience for feature weights and struggle to adapt to different charging circuit characteristics.

[0093] Example 2 like Figure 2 As shown, based on the same inventive concept as the above embodiments, the present invention also provides a centralized carport charging safety management and control device, comprising: The data acquisition module is used to acquire electrical parameter data for each charging circuit. The electrical parameter data includes at least voltage, current, residual current and temperature. The data processing module is used to preprocess electrical parameter data to obtain a multi-dimensional feature vector for each charging circuit. The multi-dimensional feature vector includes at least active power characteristics, power factor characteristics, current harmonic characteristics, temperature rise characteristics, and residual current characteristics. The status recognition module is used to identify the current charging status of the charging circuit based on multi-dimensional feature vectors. The charging status includes standby status, charging in progress status and abnormal status. The risk identification module is used to calculate the comprehensive risk value of the current charging circuit based on the charging status of the current charging circuit and multi-dimensional feature vectors, using a preset nonlinear fusion model. The instruction generation module is used to compare the comprehensive risk value with multiple preset risk level thresholds, and execute safety handling actions corresponding to the risk level based on the comparison results.

[0094] Example 3 like Figure 3 As shown, the present invention also provides an electronic device 100 for implementing a centralized carport charging safety management method; The electronic device 100 includes a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on at least one processor 102, and at least one communication bus 104.

[0095] The memory 101 can be used to store computer program 103. The processor 102 implements the steps of the centralized carport charging safety management method of Embodiment 1 by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101.

[0096] The memory 101 may primarily include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created based on the use of the electronic device 100 (such as audio data), etc. In addition, the memory 101 may include non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other non-volatile solid-state storage device.

[0097] At least one processor 102 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Processor 102 may be a microprocessor or any conventional processor. Processor 102 is the control center of electronic device 100, connecting various parts of electronic device 100 via various interfaces and lines.

[0098] The memory 101 in the electronic device 100 stores multiple instructions to implement a centralized carport charging safety management method, and the processor 102 can execute multiple instructions to achieve the following: Obtain electrical parameter data for each charging circuit, including at least voltage, current, residual current, and temperature; The electrical parameter data is preprocessed to obtain a multi-dimensional feature vector for each charging circuit. The multi-dimensional feature vector includes at least active power characteristics, power factor characteristics, current harmonic characteristics, temperature rise characteristics, and residual current characteristics. The charging state of the current charging circuit is identified based on multi-dimensional feature vectors. The charging state includes standby state, charging in progress state, and abnormal state. Based on the current charging state of the charging circuit and the multi-dimensional feature vector, the comprehensive risk value of the current charging circuit is calculated based on a preset nonlinear fusion model. The overall risk value is compared with multiple preset risk level thresholds, and safety measures corresponding to the risk level are executed based on the comparison results.

[0099] Example 4 If the modules / units integrated in the electronic device 100 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, and read-only memory (ROM).

[0100] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0101] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0102] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0103] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0104] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0105] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A centralized vehicle shed charging safety management and control method, characterized in that, include: Obtain electrical parameter data for each charging circuit. The electrical parameter data should include at least the circuit voltage, circuit current, residual current, and temperature. The electrical parameter data is preprocessed to obtain a multi-dimensional feature vector for each charging circuit. The multi-dimensional feature vector includes at least active power characteristics, power factor characteristics, current harmonic characteristics, temperature rise characteristics, and residual current characteristics. Identify the charging state of the current charging circuit based on multi-dimensional feature vectors; Based on the current charging state of the charging circuit and the multi-dimensional feature vector, the comprehensive risk value of the current charging circuit is calculated based on a preset nonlinear fusion model. The overall risk value is compared with multiple preset risk level thresholds, and safety measures corresponding to the risk level are executed based on the comparison results.

2. The centralized vehicle shed charging safety management method according to claim 1, characterized in that, Identifying the current charging state of the charging circuit based on multi-dimensional feature vectors includes the following steps: A finite state machine model is established using multidimensional feature vectors as input. The state set of the finite state machine model includes standby state, power-on state, constant current charging state, constant voltage charging state, end-of-life or fully charged state, and abnormal state. Obtain the set of configurable parameters issued by the platform. The set of configurable parameters includes at least the following: power-on confirmation delay window, constant current confirmation delay window, constant voltage confirmation delay window, end-of-charge confirmation delay window, anomaly confirmation delay window, upper and lower thresholds for the proportion of each harmonic, power change rate threshold, charging end power threshold, power factor change rate threshold, power decrease rate threshold, temperature difference change rate threshold, harmonic proportion sudden change threshold, temperature rise rate threshold, and remaining current threshold. Based on multidimensional feature vectors, configurable parameter sets, and preset state transition rules, the charging state and verification mark of the current time window are determined.

3. The centralized vehicle shed charging safety management method according to claim 2, characterized in that, State transition rules should include at least the following: When the device is in standby mode and the charging state gate signal is continuously held for a number of time windows equal to the power-on confirmation delay window, it will transition to the power-on state. When the state is in the power-on state and the proportion of each harmonic is between the corresponding upper and lower thresholds, and the absolute value of the first difference of the active power characteristic is not greater than the power change rate threshold, the state is continuously maintained for a number of time windows of constant current confirmation delay window, and then it is transferred to the constant current charging state. When the constant current charging state is maintained for a number of time windows, and the first difference of the power factor characteristic is not less than the power factor change rate threshold, and the first difference of the active power characteristic is not greater than the negative power decrease rate threshold, the constant voltage confirmation delay window state is maintained for a number of time windows. When the state of constant voltage charging, active power characteristic not greater than the charging end power threshold, and absolute value of the second difference of temperature difference characteristic not greater than the temperature difference change rate threshold is maintained for a number of time windows for the closing confirmation delay window, the state is transferred to the closing or full charge state. When the maximum value of the first-order difference of the harmonic proportion characteristic is not less than the harmonic proportion mutation threshold, or the temperature rise rate characteristic is not less than the temperature rise rate threshold, or the residual current characteristic is not less than the residual current threshold, and the state is continuously maintained for the number of time windows of the abnormal confirmation delay window, the state is transferred to the abnormal state and a verification mark is set.

4. The centralized vehicle shed charging safety management method according to claim 3, characterized in that, Overall Risk Value The calculation method is as follows: in, Configurable weighting coefficients for state factors; This is a state factor used to quantify the contribution of the current charging state to the risk. The current charging state of the charging circuit ρ; Configurable weight coefficients for the i-th type of feature; For the nonlinear mapping function of the i-th type of feature, respectively, it corresponds to the quantification of the degree of anomaly of active power feature, power factor feature, current harmonic feature, temperature rise feature and residual current feature; For the charging circuit ρ within the time window The multidimensional feature vector.

5. The centralized vehicle shed charging safety management method according to claim 4, characterized in that, The overall risk value is compared with multiple preset risk level thresholds, and safety procedures corresponding to the risk level are executed based on the comparison results, including: Obtain the hierarchical parameter set issued by the platform, which includes at least: a first risk threshold, a second risk threshold, a third risk threshold, and a first-level alarm delay window, a second-level alarm delay window, and a third-level alarm delay window; wherein the first risk threshold is less than the second risk threshold, and the second risk threshold is less than the third risk threshold; Three levels of risk conditions are defined; among them, the severe risk condition is: the comprehensive risk value is greater than the third risk threshold, or the charging state is abnormal; the medium risk condition is: the comprehensive risk value is greater than the second risk threshold, or the review mark is true; the warning risk condition is: the current charging state is and the comprehensive risk value is greater than the first risk threshold. The application of a delay-hold operator determines the three levels of risk conditions. Specifically, if a severe risk condition is maintained for a number of consecutive time windows corresponding to the third-level alarm delay window, then the third-level alarm level is output; if a moderate risk condition is maintained for a number of consecutive time windows corresponding to the second-level alarm delay window, then the second-level alarm level is output; if a warning risk condition is maintained for a number of consecutive time windows corresponding to the first-level alarm delay window, then the first-level alarm level is output; otherwise, no alarm level is output. The corresponding safety procedures are executed according to the alarm level. Specifically, when the alarm level is Level 1, a prompting action is executed, including at least an audible and visual alarm; when the alarm level is Level 2, a control action is executed, including at least stopping charging or limiting the charging current; and when the alarm level is Level 3, a power-off action is executed, including at least disconnecting the power supply relay or contactor.

6. The centralized vehicle shed charging safety management method according to claim 5, characterized in that, Also includes: The comprehensive risk value, corresponding risk level, and corresponding multi-dimensional feature vector of the current charging circuit are encapsulated into a unified event frame and uploaded to the remote management platform. The remote management platform extracts multidimensional feature vectors and corresponding risk levels from multiple received event frames to construct a labeled sample set; Using the sample set as input, the configurable parameters in the nonlinear fusion model are iteratively optimized by minimizing the cross-entropy loss function with a regularization term; the optimized configurable parameters are then sent to the front end to update subsequent risk calculations.

7. The centralized vehicle shed charging safety management method according to claim 6, characterized in that, Iterative optimization is performed on the configurable parameters in the nonlinear fusion model; the optimization problem is: Where θ represents all parameters to be optimized; σ( () is the Sigmoid function; Let be the overall risk value of the i-th sample; The corresponding risk level threshold; This represents the true risk label for the i-th sample. The total number of samples; This is the regularization coefficient.

8. A centralized vehicle shed charging safety management and control device, characterized in that, include: The data acquisition module is used to acquire electrical parameter data for each charging circuit. The electrical parameter data includes at least circuit voltage, circuit current, residual current, and temperature. The data processing module is used to preprocess electrical parameter data to obtain a multi-dimensional feature vector for each charging circuit. The multi-dimensional feature vector includes at least active power characteristics, power factor characteristics, current harmonic characteristics, temperature rise characteristics, and residual current characteristics. The status recognition module is used to identify the current charging state of the charging circuit based on multi-dimensional feature vectors. The risk identification module is used to calculate the comprehensive risk value of the current charging circuit based on the charging status of the current charging circuit and multi-dimensional feature vectors, using a preset nonlinear fusion model. The instruction generation module is used to compare the comprehensive risk value with multiple preset risk level thresholds, and execute safety handling actions corresponding to the risk level based on the comparison results.

9. An electronic device, characterized in that, It includes a processor and a memory, the processor being used to execute a computer program stored in the memory to implement the centralized carport charging safety management method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one instruction, which, when executed by a processor, implements the centralized vehicle shed charging safety management method as described in any one of claims 1 to 7.

Images