Real-time data analysis system and method for internet of things terminals

Abstract

This invention discloses a real-time data analysis system and method for IoT terminals, relating to the field of data analysis. By collecting data on the IoT terminal's guidance voltage, charging current, control board temperature, connector temperature, and state switching events, the system determines the rising zero-crossing moment, divides the control guidance cycle, and calculates the cycle duration and number of samples to form a control guidance cycle sequence, aligning the aforementioned data with the state switching events. For each state switching moment, a pre-switching window and a post-switching window are constructed, extracting the number of samples, duty cycle, and average current and temperature values ​​within the window, and integrating them into a window feature package. Four Boolean values ​​are obtained by examining whether the duty cycle remains unchanged, whether the cycle changes, and the differences in current and temperature changes. Logical judgments are performed based on these four Boolean values ​​to obtain attributions such as control guidance frequency drift pseudo-switching and true thermal derating, improving the efficiency of data analysis and the accuracy of attribution.

Description

Technical Field

[0001] This invention relates to the field of data analysis, and more specifically to a real-time data analysis system and method for Internet of Things (IoT) terminals. Background Technology

[0002] As a core IoT terminal supporting vehicle recharging, the operational stability and status traceability of intelligent electric vehicle charging piles directly affect user charging experience, equipment safety, and operational efficiency. These charging piles achieve command interaction and charging process control between the vehicle and the charging pile through control guidance circuit signals. The charging pile outputs a ±12V pulse width modulated signal at a nominal frequency of 1kHz to the vehicle. The duty cycle of the control guidance circuit signal is used to encode the allowed charging current value, while the frequency itself only serves as a signal synchronization reference and is not directly related to charging permissions. Simultaneously, the charging pile needs to collect key parameters such as charging current, control board temperature, and connector temperature in real time and upload its operational status to the operation and maintenance platform. In actual operation and maintenance, the platform needs to use this data to trace the root cause of any status change to identify the responsible party and ensure the continuity of charging services.

[0003] Under high-load operation scenarios, the charging pile control board experiences clock source thermal drift due to increased heat dissipation load from the power module. This causes the control guidance circuit signal frequency to deviate from the nominal value of 1kHz. When the vehicle-side control guidance circuit signal detection module detects the frequency abnormality, it may mistakenly determine that the signal is illegal and trigger an instantaneous operation state switch. However, this state switch is a pseudo-switching, not a genuine operation or overheat protection-induced operation state switch. Essentially, it is a vehicle-side misjudgment caused by control guidance frequency drift. Existing systems cannot effectively distinguish between such pseudo-switching and genuine switching caused by control guidance frequency drift, leading to problems such as confusion in liability determination, broken evidence chains in compliance audits, and interference with billing accuracy. This seriously affects the operation and maintenance efficiency and service quality of charging piles. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention proposes a real-time data analysis system and method for IoT terminals, which solves the problem that they cannot effectively distinguish between pseudo-switching and real switching of control guidance frequency drift.

[0005] To achieve the above objectives, the present invention provides the following technical solution: The system collects control guidance circuit signals that represent the command interaction between the charging pile and the vehicle, and obtains their real-time voltage values ​​according to the sampling frequency. The real-time voltage values ​​are marked as guidance voltages. When the guidance voltage rises from a negative voltage to a non-negative voltage, the sampling time of the non-negative voltage is represented as the rising zero-crossing time. The rising zero-crossing time is represented as the starting point of a control guidance cycle. The control guidance cycle is the time interval between two adjacent rising zero-crossing times in the control guidance circuit signals output by the charging pile. The number of samples is obtained by multiplying the difference between two adjacent rising zero crossing moments by the sampling frequency. The number of high-level samples is obtained by determining whether the conduction voltage of each sample is greater than zero. The duty cycle is obtained by dividing the number of high-level samples by the number of samples. Acquire charging current, control board temperature and connector temperature; acquire state transition events, including target operating state and state transition time. Construct a pre-switching window and a post-switching window based on the time intervals between two adjacent rising zero-crossing moments when the state switching time occurs; The number of samples, the number of high-level samples, the duty cycle, the charging current, the control board temperature, the connector temperature, and the corresponding state switching time for the window before and after switching are integrated into a window feature package.

[0006] Furthermore, the sampling time of the acquisition of the pilot voltage is recorded, and the pilot voltage at the (i-1)th sampling time is obtained. If the pilot voltage at the (i-1)th sampling time is negative and the pilot voltage at the ith sampling time is non-negative, then the ith sampling time is determined to be the rising zero-crossing time. All rising zero-crossing times are obtained and sorted in chronological order.

[0007] Furthermore, the duration of the i-th control guidance cycle is obtained by subtracting the i-th rising zero-crossing time from the (i+1)-th rising zero-crossing time. The number of samples in the i-th control guidance cycle is obtained by multiplying the acquisition frequency by the cycle duration of the i-th control guidance cycle. The sampled samples include the guidance voltage. The number of samples that satisfy the guide voltage > 0 is counted and marked as the number of high-level samples.

[0008] Furthermore, the rising zero-crossing moments, the number of sampled samples, the number of high-level samples, and the corresponding duty cycles within all control and guidance cycles are combined into a control and guidance cycle sequence; Based on the rising zero-crossing moment of the control guidance cycle sequence, the charging current, control board temperature, and connector temperature are assigned to their respective control guidance cycle sequences, and the state switching events are associated with the corresponding control guidance cycle sequences.

[0009] Furthermore, for each state transition moment, the rising zero-crossing moments in the control guidance cycle sequence are traversed. The control guidance cycle with the i-th rising zero-crossing moment ≤ state transition moment < i+1 rising zero-crossing moment is taken as the pre-switching window, and the control guidance cycle from the i+1 rising zero-crossing moment to the i+2 rising zero-crossing moment is taken as the post-switching window.

[0010] Furthermore, the number of sampled samples, the number of high-level samples, and the duty cycle of the control guidance cycle sequence corresponding to the window before and after switching are extracted respectively. Calculate the average value of the charging current, the average value of the control board temperature, and the average value of the connector temperature for the control guidance cycle sequence corresponding to the window before and after switching. The number of samples, number of high-level samples, duty cycle, average charging current, average control board temperature, average connector temperature, and the i-th state switching time corresponding to the pre-switching window and post-switching window are integrated into a window feature packet, and the window feature packet corresponding to the state switching time in each state switching event is output. The first Boolean value, the second Boolean value, the third Boolean value, and the fourth Boolean value are obtained by analyzing and judging the window feature packet; Attribution judgments are made based on the first, second, third, and fourth Boolean values, and an attribution evidence package is generated based on the attribution judgment results.

[0011] Furthermore, an equality check is performed based on the number of high-level samples and the number of sampled samples in the window feature packet. The first Boolean value is then marked according to the check result, as follows: If the number of high-level samples in the window before switching × the number of samples in the window after switching = the number of high-level samples in the window after switching × the number of samples in the window before switching, then mark the first Boolean value as 1; otherwise, mark the first Boolean value as 0. Inequality checks are performed based on the number of samples in the window feature packet, and the second Boolean value is marked according to the check result, as follows: If the number of samples in the window before switching is not equal to the number of samples in the window after switching, then mark the second Boolean value as 1; otherwise, mark the second Boolean value as 0.

[0012] Furthermore, the average charging current of the window before switching is subtracted from the average charging current of the window after switching in the window feature packet to obtain the current difference. The verification is performed based on the value of the current difference, and the third Boolean value is marked according to the verification result, as follows: If the current difference is <0, mark the third Boolean value as 1; if the current difference is ≥0, mark the third Boolean value as 0. The change in control panel temperature is obtained by subtracting the average control panel temperature of the window before switching from the average control panel temperature of the window after switching in the window feature package. The change in connector temperature is obtained by subtracting the average connector temperature of the window before switching from the average connector temperature of the window after switching in the window feature package. Verification was performed by comparing the relative magnitudes of the temperature changes on the control board and the connectors. Based on the verification results, a fourth Boolean value was assigned, as follows: If the temperature change of the control board is greater than the temperature change of the connector, mark the fourth Boolean value as 1; if the temperature change of the control board is less than or equal to the temperature change of the connector, mark the fourth Boolean value as 0.

[0013] Furthermore, attribution judgments are performed based on the first, second, third, and fourth Boolean values ​​to obtain the judgment results, as follows: If the first Boolean value, the second Boolean value, the third Boolean value, and the fourth Boolean value are all 1 at the same time, then the target's operating state is determined to be a false switch of control guidance frequency drift. If the first Boolean value is 0 and the third Boolean value is 1, then the target operating state is determined to be either true thermal derating or over-temperature protection. The judgment result is associated with the corresponding state transition event to generate an attribution evidence package, which includes the judgment result and the corresponding state transition event.

[0014] Furthermore, a real-time data analysis system for IoT terminals is proposed to implement any of the real-time data analysis methods described above, including: The data acquisition module collects the guidance voltage, charging current, control board temperature and connector temperature, and obtains state switching events. It determines the rising zero crossover moment, calculates the control guidance cycle duration and the number of samples, forms the control guidance cycle sequence, and then aligns the data with the state switching events. The window construction module extracts the state switching time based on the state switching event, constructs a pre-switching window and a post-switching window for each state switching time, extracts data related to the control and guidance cycle sequence within the window, and integrates them into a window feature package for the corresponding state switching time. The signal analysis module uses integer equations to check whether the duty cycle of the window before and after switching remains unchanged, thus obtaining the first Boolean value; and uses integer inequalities to check whether the period changes, thus obtaining the second Boolean value. The electrothermal analysis module obtains the third Boolean quantity by the sign of the current difference, and obtains the fourth Boolean quantity by comparing the temperature change of the control board with that of the connector. The attribution discrimination module generates and outputs an attribution evidence package based on the logical values ​​of four Boolean values, correlates with state switching events, and then uses this information to determine the attribution evidence.

[0015] Compared with existing technologies, it has the following advantages: This solution proposes a real-time data analysis system and method for IoT terminals. By using the rising zero-crossing moment of the control guidance cycle as the data alignment benchmark, it effectively solves the problem of cycle misalignment that easily occurs when using conventional system timestamp alignment in existing technologies, significantly improving the targeting of data correlation in abnormal scenarios. This solution accurately locates the starting point of each control guidance cycle by capturing the level change of the control guidance circuit signal from a negative voltage at the previous moment to a non-negative voltage at the current moment. Based on this, charging current, control board temperature, connector temperature, and state switching events are divided into corresponding control guidance cycle sequences. This ensures that subsequent analysis can retrieve data within the same cycle dimension without additional time matching, avoiding the problem of irrelevant time dimension data interfering with the analysis results, and laying the foundation for accurate analysis of control guidance circuit signal anomalies.

[0016] By constructing pre-switching and post-switching windows, precise filtering of state-switching related data is achieved, overcoming the shortcomings of existing technologies that use fixed-length windows, which easily introduce irrelevant periodic data. This scheme selects the complete control guidance cycle in which the state switch occurs as the pre-switching window and the immediately following control guidance cycle as the post-switching window. This preserves the integrity of the control guidance cycle while focusing the analysis on the key cycle closest to the state switch. Simultaneously, by integrating the number of sampled samples, the number of high-level samples, the duty cycle, and the average current and temperature values ​​within the window through window feature packets, it achieves one-stop integration of control guidance circuit signals, power supply, heat source, and operating status. This significantly improves data analysis efficiency and data transmission consistency, providing a unified and complete input interface for subsequent attribution.

[0017] By replacing floating-point arithmetic with integer operations to verify the control and guidance circuit signals, calculation errors are effectively avoided, and signal analysis accuracy is improved. Existing technologies that directly compare floating-point duty cycles are prone to misjudgment due to precision loss. This solution uses integer equations to verify whether the duty cycle remains unchanged and integer inequalities to verify whether the control and guidance period changes, outputting the first and second Boolean values ​​respectively to accurately anchor the core of the control and guidance frequency drift.

[0018] By using electrothermal analysis based on physical mechanisms and four Boolean quantities as a benchmark, this invention effectively distinguishes between false handovers and genuine faults. It can accurately identify false handovers due to control guidance frequency drift, genuine thermal derating or over-temperature protection, and other events, avoiding biased liability determination caused by confusion between false handovers and genuine protection in logs. The attribution evidence package also provides clear and traceable conclusions for operation and maintenance audits, effectively improving problem location efficiency, reducing billing disputes and unnecessary equipment maintenance, and ensuring the stability and compliance of IoT terminal operations. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2 This is a schematic diagram of the system framework of the present invention; Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Please see Figure 1 This application provides a real-time data analysis method for IoT terminals; As an embodiment of this application, the method specifically includes: The system collects control guidance circuit signals that represent the command interaction between the charging pile and the vehicle, and obtains their real-time voltage values ​​according to the sampling frequency. The real-time voltage values ​​are marked as guidance voltages. When the guidance voltage rises from a negative voltage to a non-negative voltage, the sampling time of the non-negative voltage is represented as the rising zero-crossing time. The rising zero-crossing time is represented as the starting point of a control guidance cycle. The control guidance cycle is the time interval between two adjacent rising zero-crossing times in the control guidance circuit signals output by the charging pile. The number of samples is obtained by multiplying the difference between two adjacent rising zero crossing moments by the sampling frequency. The number of high-level samples is obtained by determining whether the conduction voltage of each sample is greater than zero. The duty cycle is obtained by dividing the number of high-level samples by the number of samples. Acquire charging current, control board temperature and connector temperature; acquire state transition events, including target operating state and state transition time. Construct a pre-switching window and a post-switching window based on the time intervals between two adjacent rising zero-crossing moments when the state switching time occurs; The number of samples, the number of high-level samples, the duty cycle, the charging current, the control board temperature, the connector temperature, and the corresponding state switching time for the window before and after switching are integrated into a window feature package.

[0022] As a second embodiment of this application, this embodiment is implemented based on the first embodiment, and the method provided in this embodiment includes the following steps: Step 1: Collect the guiding voltage through an IoT terminal and record the sampling time. In this example, the IoT terminal is a smart electric vehicle charging pile. Iterate through all sampling times of the guiding voltage. For the i-th sampling time, obtain the guiding voltage at the (i-1)-th sampling time. If the guiding voltage at the (i-1)-th sampling time is negative and the guiding voltage at the i-th sampling time is non-negative, then the i-th sampling time is determined to be a rising zero-crossing moment. Obtain all rising zero-crossing moments and sort them in chronological order. Specifically, controlling the guiding circuit signal is a key technology in the electric vehicle charging system. The guiding circuit is the transmission carrier of the control signal, and the control signal is the core function of this circuit, used to realize the connection between the charging pile and the vehicle. In this example, the control guidance circuit signal for vehicle-to-vehicle command interaction and charging control is designed as a ±12V pulse width modulation signal wave. Negative voltage is a component of the signal level, alternating with positive voltage to form a complete waveform to realize the command interaction and charging control between the charging pile and the vehicle. The rising zero crossover moment is the starting point of a control guidance cycle. The control guidance cycle is the time interval between two adjacent rising zero crossover moments in the control guidance circuit signal output by the charging pile, that is, the time for the control guidance circuit signal to complete one high-level and low-level cycle. By the level change of negative in the previous moment and non-negative in the current moment, the starting point of each control guidance cycle is accurately captured, providing a unique time dimension for the subsequent control guidance cycle division and avoiding cycle misalignment across moments. For the i-th rising zero-crossing moment, subtracting the i-th rising zero-crossing moment from the (i+1)-th rising zero-crossing moment gives the duration of the i-th control guidance cycle. Multiplying the sampling frequency of the IoT terminal for collecting the guidance voltage by the duration of the cycle and rounding the product gives the number of samples in a control guidance cycle. Traverse the sampling samples of the i-th control guidance cycle, determine whether the guidance voltage of each sampling sample is greater than zero, count the number of all sampling samples that satisfy guidance voltage > 0, and mark this number as the number of high-level samples. The number of high-level samples can reflect the proportion of high level in the control guidance cycle and is the core parameter for calculating the duty cycle. Divide the number of high-level samples by the number of sampled samples to obtain the duty cycle of the control guidance period; The rising zero-crossing moments, number of sampled samples, number of high-level samples, and duty cycle within all control and guidance cycles are combined into a control and guidance cycle sequence. The system acquires the charging current, control board temperature (reflecting the thermal state of the internal circuitry), and connector temperature (reflecting the contact thermal state of the charging gun head) collected by the IoT terminal. It also acquires state switching events, which include the target operating state after the IoT terminal's operating state is switched (such as charging, paused, or faulty) and records the time point when the operating state changes, i.e., the state switching moment. Based on the rising zero-crossing moment of the control guidance cycle sequence, the charging current, control board temperature, and connector temperature are assigned to their respective control guidance cycle sequences, and the state switching events are associated with the corresponding control guidance cycle sequences. Specifically, using the control and guidance cycle sequence as the alignment benchmark, unlike conventional system timestamp alignment, it focuses on scenarios where control and guidance circuit signals are abnormal, improving the targeting of subsequent analysis and ensuring that subsequent steps can call the data associated with the control and guidance cycle within the same cycle dimension without the need for additional time matching.

[0023] Step 2: Iterate through the state transition events, enumerating and extracting each state transition moment one by one. Combine all the extracted state transition moments into a set of state transition moments. For each state transition moment in the set of state transition moments, construct a window before the transition and a window after the transition. The specific construction process is as follows: For each state transition time, iterate through the rising zero-crossing times in the control guidance cycle sequence, and take the control guidance cycle with the i-th rising zero-crossing time ≤ state transition time < i+1 rising zero-crossing time as the pre-switching window, and take the control guidance cycle from the i+1 rising zero-crossing time to the i+2 rising zero-crossing time as the post-switching window. Specifically, by selecting a complete control guidance cycle in which the state transition occurs as the pre-transition window and the next control guidance cycle immediately following this as the post-transition window, the integrity of the control guidance cycle is preserved. This approach also distinguishes the analysis from conventional fixed-length windows, narrowing the analysis scope to the time window closest to the state transition and avoiding the introduction of more irrelevant data, thus making the subsequent analysis of the causes of the state transition more accurate. Extract the number of sampled samples, the number of high-level samples, and the duty cycle of the control guidance cycle sequence corresponding to the window before and after switching, respectively; Calculate the average value of the charging current, the average value of the control board temperature, and the average value of the connector temperature for the control guidance cycle sequence corresponding to the window before and after switching. The number of samples, number of high-level samples, duty cycle, average charging current, average control board temperature, average connector temperature, and the i-th state switching time corresponding to the pre-switching window and post-switching window are integrated into a window feature packet, and the window feature packet corresponding to the state switching time in each state switching event is output. Specifically, by generating window feature packets corresponding to state transition moments, a one-stop integration of control guidance circuit signals, power supply, heat source, and operating status is achieved, enabling subsequent steps to directly call all data within the packet, thus improving data analysis efficiency. The structured design of the window feature packets ensures the integrity and consistency of data transmission, providing a unified input interface for deterministic attribution.

[0024] Step 3: Based on the duty cycle calculation method (number of high-level samples divided by the number of sampled samples), if the duty cycle of the window remains unchanged before and after the state transition, it must satisfy the condition that the duty cycle of the window before the transition in the window feature packet equals the duty cycle of the window after the transition. That is, the number of high-level samples of the window before the transition divided by the number of sampled samples equals the number of high-level samples of the window after the transition divided by the number of sampled samples. To avoid floating-point arithmetic errors, this is converted to an integer equation for verification. The number of high-level samples in the window before switching × the number of samples in the window after switching = the number of high-level samples in the window after switching × the number of samples in the window before switching; If the equation is true, mark the first Boolean value as 1, indicating that the duty cycle remains unchanged and the IoT terminal has not actively adjusted the business semantics. If the equation is false, mark the first Boolean value as 0. Specifically, the use of integer cross-multiplication equations to replace direct comparison of floating-point duty cycles ensures calculation accuracy. The core logical starting point for distinguishing between pseudo-switching of control guidance frequency drift (duty cycle unchanged) and true thermal derating (duty cycle change) is to directly verify whether the duty cycle remains unchanged by marking the first Boolean value. In this example, the nominal frequency of the control guidance circuit signal is 1kHz, and its period is determined by the number of samples and the acquisition frequency. If the control guidance period changes, i.e., the frequency deviates from 1kHz, the number of samples in the window before and after the switch will inevitably be different. Therefore, the following verification is required: The number of samples sampled in the window before switching is not the same as the number of samples sampled in the window after switching. If the inequality is true, mark the second Boolean value as 1, indicating that the control guidance period changes and the frequency deviates by 1kHz; if the inequality is false, mark the second Boolean value as 0. Specifically, the core logical basis for identifying false switching (frequency change) of control guidance frequency drift is to directly verify whether the period frequency has changed based on the integer inequality test. This step, based on integer equations and integer inequalities, completes the test for the constant duty cycle and the change of period of the control guidance circuit signal. The first and second Boolean values ​​output directly anchor the core characteristics of false switching of control guidance circuit signal frequency drift, which is the key intermediate conclusion for subsequent deterministic attribution.

[0025] Step 4: Calculate the current difference by subtracting the average charging current of the window before the switch from the average charging current of the window after the switch in the window feature packet. Use the sign of the current difference to verify the change in power supply status. If the current difference is <0, mark the third Boolean value = 1, indicating that the average current is less than before the switch after switching to the target operating state. That is, the switching of the operating state is accompanied by a decrease or interruption of the charging current. If the current difference is ≥0, it means that the current has not decreased, and mark the third Boolean value = 0. Specifically, judging the current change by the sign of the current difference can directly verify whether the state switch is accompanied by a decrease in power supply. This is the key basis for distinguishing between false switching (current decrease caused by vehicle-side misjudgment) and non-power supply events. The control board temperature change is obtained by subtracting the average control board temperature of the window before the switch from the average control board temperature of the window after the switch in the window feature package. Similarly, the connector temperature change is obtained by subtracting the average connector temperature of the window before the switch from the average connector temperature of the window after the switch in the window feature package. The relative magnitudes of the control board temperature change and the connector temperature change are then compared for verification. If the temperature change of the control board is greater than the temperature change of the connector, mark the fourth Boolean value as 1. This indicates that the temperature rise of the control board is greater than that of the connector, meaning that the temperature rise is more biased towards the control board side. This is consistent with the physical mechanism of control guidance frequency drift caused by thermal drift of the IoT terminal clock source. If the temperature change of the control board is less than or equal to the temperature change of the connector, mark the fourth Boolean value as 0. This indicates that the temperature rise is more biased towards the connector side. This is consistent with the common scenario where real thermal derating or over-temperature protection is triggered by connector overheating. Specifically, based on the physical mechanism of the control board temperature rise being more significant due to clock source thermal drift, this provides scenario-level verification for pseudo-switching of control guide frequency drift, distinguishing it from the more significant connector temperature rise caused by real thermal protection. By correlating the physical mechanism of control guide clock source thermal drift with the relative magnitude of temperature changes in the control board and connector, the abstract physical cause is transformed into a quantifiable and comparable temperature difference relationship, providing a creative verification dimension at the physical scenario level for attributing pseudo-switching, complementing the verification of pseudo-switching of control guide circuit signal frequency drift.

[0026] Step 5: Obtain the first, second, third, and fourth Boolean values. Based on these four Boolean values, perform logical judgments to determine the cause of control guidance frequency drift pseudo-switching, actual thermal derating or over-temperature protection, and other events. Output an evidence package for operation and maintenance auditing, as follows: If the first, second, third, and fourth Boolean values ​​are all 1, then the current state switch of the IoT terminal is determined to be a false switch due to control guidance frequency drift. Specifically, the control guidance circuit signal transmitted from the smart charging pile (IoT terminal) to the vehicle is drifted due to the temperature of the IoT terminal's clock source caused by high temperature, causing the frequency of the control guidance circuit signal to deviate from the standard 1kHz, but the duty cycle still maintains the charging pile's preset value. At this time, the vehicle side misjudges the signal as illegal because the control guidance frequency does not meet the standard, and then briefly switches the charging state, such as switching from charging to charging pause. However, this state switch is not caused by real thermal derating or over-temperature protection. Real protection will actively adjust the duty cycle or cut off the power. It is a false state switch, that is, a false switch due to control guidance frequency drift. Through multi-dimensional verification of the control guidance circuit signal and the real physical scenario, the accurate identification of false switches due to control guidance frequency drift has been achieved. If the first Boolean value is 0 and the third Boolean value is 1, it is determined to be a true thermal derating or over-temperature protection, indicating that the IoT terminal achieves thermal protection by adjusting the duty cycle. If the above two conditions are not met, it is determined to be an electrical fault, contact problem, or user operation, and further investigation is required; The judgment results are associated with the corresponding state transition events to form an attribution evidence package. Specifically, the structured output of the attribution evidence package enables different roles to quickly obtain key information, improves the efficiency of problem localization, and provides clear and traceable conclusions for IoT terminal operation and maintenance audits.

[0027] Furthermore, refer to Figure 2 As shown, a real-time data analysis system for IoT terminals is proposed to implement any of the real-time data analysis methods described above, including: The data acquisition module collects voltage, charging current, control board temperature and connector temperature, and obtains state switching events. It determines the rising zero crossover moment, calculates the control guidance cycle duration and the number of samples, forms a control guidance cycle sequence, and then aligns the data with the state switching events. The window construction module extracts the state switching time based on the state switching event, constructs a pre-switching window and a post-switching window for each state switching time, extracts data related to the control and guidance cycle sequence within the window, and integrates them into a window feature package for the corresponding state switching time. The signal analysis module uses integer equations to check whether the duty cycle of the window before and after switching remains unchanged, thus obtaining the first Boolean value; and uses integer inequalities to check whether the period changes, thus obtaining the second Boolean value. The electrothermal analysis module obtains the third Boolean quantity by the sign of the current difference, and obtains the fourth Boolean quantity by comparing the temperature change of the control board with that of the connector. The attribution discrimination module generates and outputs an attribution evidence package based on the logical values ​​of four Boolean values, correlates with state switching events, and then uses this information to determine the attribution evidence.

[0028] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A real-time data analysis method for IoT terminals, characterized in that, include: The system collects control guidance circuit signals that represent the command interaction between the charging pile and the vehicle, and obtains their real-time voltage values ​​according to the sampling frequency. The real-time voltage values ​​are marked as guidance voltages. When the guidance voltage rises from a negative voltage to a non-negative voltage, the sampling time of the non-negative voltage is represented as the rising zero-crossing time. The rising zero-crossing time is represented as the starting point of a control guidance cycle. The control guidance cycle is the time interval between two adjacent rising zero-crossing times in the control guidance circuit signals output by the charging pile. The number of samples is obtained by multiplying the difference between two adjacent rising zero crossing moments by the sampling frequency. The number of high-level samples is obtained by determining whether the conduction voltage of each sample is greater than zero. The duty cycle is obtained by dividing the number of high-level samples by the number of samples. Acquire charging current, control board temperature and connector temperature; acquire state transition events, including target operating state and state transition time. Construct a pre-switching window and a post-switching window based on the time intervals between two adjacent rising zero-crossing moments when the state switching time occurs; The number of samples, the number of high-level samples, the duty cycle, the charging current, the control board temperature and the connector temperature, as well as the corresponding state switching time, corresponding to the window before and after the switching are integrated into a window feature package. The specific methods for constructing the window before and after the switch include: For each state transition time, iterate through the rising zero-crossing times in the control guidance cycle sequence, and take the control guidance cycle with the i-th rising zero-crossing time ≤ state transition time < i+1 rising zero-crossing time as the pre-switching window, and take the control guidance cycle from the i+1 rising zero-crossing time to the i+2 rising zero-crossing time as the post-switching window. Extract the number of sampled samples, the number of high-level samples, and the duty cycle of the control guidance cycle sequence corresponding to the window before and after switching, respectively; Calculate the average value of the charging current, the average value of the control board temperature, and the average value of the connector temperature for the control guidance cycle sequence corresponding to the window before and after switching. The number of samples, number of high-level samples, duty cycle, average charging current, average control board temperature, average connector temperature, and the i-th state switching time corresponding to the pre-switching window and post-switching window are integrated into a window feature packet, and the window feature packet corresponding to the state switching time in each state switching event is output. The first Boolean value, the second Boolean value, the third Boolean value, and the fourth Boolean value are obtained by analyzing and judging the window feature packet; Attribution judgments are made based on the first, second, third, and fourth Boolean values, and an attribution evidence package is generated based on the attribution judgment results.

2. The real-time data analysis method for IoT terminals according to claim 1, characterized in that, Methods for determining the rising zero crossover moment include: Record the sampling time of the acquisition of the pilot voltage, obtain the pilot voltage at the (i-1)th sampling time. If the pilot voltage at the (i-1)th sampling time is negative and the pilot voltage at the ith sampling time is non-negative, then the ith sampling time is determined to be the rising zero-crossing time. Obtain all rising zero-crossing times and sort them in chronological order.

3. The real-time data analysis method for IoT terminals according to claim 1, characterized in that, Methods for obtaining the number of high-level samples include: Subtracting the i-th rising zero crossover time from the (i+1)-th rising zero crossover time gives the duration of the i-th control guidance cycle. The number of samples in the i-th control guidance cycle is obtained by multiplying the acquisition frequency by the cycle duration of the i-th control guidance cycle. The sampled samples include the guidance voltage. The number of samples that satisfy the guide voltage > 0 is counted and marked as the number of high-level samples.

4. The real-time data analysis method for IoT terminals according to claim 1, characterized in that, include: The rising zero-crossing moments, number of sampled samples, number of high-level samples, and corresponding duty cycles within all control and guidance cycles are combined into a control and guidance cycle sequence. Based on the rising zero-crossing moment of the control guidance cycle sequence, the charging current, control board temperature, and connector temperature are assigned to their respective control guidance cycle sequences, and the state switching events are associated with the corresponding control guidance cycle sequences.

5. The real-time data analysis method for IoT terminals according to claim 1, characterized in that, The methods for obtaining the first and second Boolean values ​​include: An equality check is performed based on the number of high-level samples and the number of sampled samples in the window feature packet. The first Boolean value is then marked according to the check result, as follows: If the number of high-level samples in the window before switching × the number of samples in the window after switching = the number of high-level samples in the window after switching × the number of samples in the window before switching, then mark the first Boolean value as 1; otherwise, mark the first Boolean value as 0. Inequality checks are performed based on the number of samples in the window feature packet, and the second Boolean value is marked according to the check result, as follows: If the number of samples in the window before switching is not equal to the number of samples in the window after switching, then mark the second Boolean value as 1; otherwise, mark the second Boolean value as 0.

6. The real-time data analysis method for IoT terminals according to claim 1, characterized in that, The methods for obtaining the third and fourth Boolean values ​​include: The current difference is obtained by subtracting the average charging current of the window before switching from the average charging current of the window after switching in the window feature packet. The verification is performed based on the value of the current difference, and the third Boolean value is marked according to the verification result, as follows: If the current difference is <0, mark the third Boolean value as 1; if the current difference is ≥0, mark the third Boolean value as 0. The change in control panel temperature is obtained by subtracting the average control panel temperature of the window before switching from the average control panel temperature of the window after switching in the window feature package. The change in connector temperature is obtained by subtracting the average connector temperature of the window before switching from the average connector temperature of the window after switching in the window feature package. Verification was performed by comparing the relative magnitudes of the temperature changes on the control board and the connectors. Based on the verification results, a fourth Boolean value was assigned, as follows: If the temperature change of the control board is greater than the temperature change of the connector, mark the fourth Boolean value as 1; if the temperature change of the control board is less than or equal to the temperature change of the connector, mark the fourth Boolean value as 0.

7. The real-time data analysis method for IoT terminals according to claim 1, characterized in that, The methods for generating the attribution evidence package include: Attribution judgments are made based on the first, second, third, and fourth Boolean values, and the judgment results are as follows: If the first Boolean value, the second Boolean value, the third Boolean value, and the fourth Boolean value are all 1 at the same time, then the target's operating state is determined to be a false switch of control guidance frequency drift. If the first Boolean value is 0 and the third Boolean value is 1, then the target operating state is determined to be either true thermal derating or over-temperature protection. The judgment result is associated with the corresponding state transition event to generate an attribution evidence package, which includes the judgment result and the corresponding state transition event.

8. A real-time data analysis system for Internet of Things (IoT) terminals, used to implement the real-time data analysis method as described in any one of claims 1-7, characterized in that, include: The data acquisition module collects the guidance voltage, charging current, control board temperature and connector temperature, and obtains state switching events. It determines the rising zero crossover moment, calculates the control guidance cycle duration and the number of samples, forms the control guidance cycle sequence, and then aligns the data with the state switching events. The window construction module extracts the state switching time based on the state switching event, constructs a pre-switching window and a post-switching window for each state switching time, extracts data related to the control and guidance cycle sequence within the window, and integrates them into a window feature package for the corresponding state switching time. The signal analysis module uses integer equations to check whether the duty cycle of the window before and after switching remains unchanged, thus obtaining the first Boolean value; and uses integer inequalities to check whether the period changes, thus obtaining the second Boolean value. The electrothermal analysis module obtains the third Boolean quantity by the sign of the current difference, and obtains the fourth Boolean quantity by comparing the temperature change of the control board with that of the connector. The attribution discrimination module generates and outputs an attribution evidence package based on the logical values ​​of four Boolean values, correlates with state switching events, and then uses this information to determine the attribution evidence.

Images