A photovoltaic inverter insulation resistance intelligent detection system and method

Abstract

The present invention discloses an intelligent detection system and method for the insulation resistance of a photovoltaic inverter, which specifically relates to the technical field of photovoltaic equipment operation safety monitoring, and includes the following steps: by collecting the bus voltage in the state transition stage and constructing a dynamic convergence model, a frozen window is generated to shield interference data; in a stable state, abnormal trends are identified based on resistance derivative changes, and a dual-path decision-making process is input in combination with environmental and grid parameters to dynamically output the operating status level; remote prompts, power reduction or output disconnection and other control strategies are executed according to the status level to achieve accurate identification and graded response control of the insulation status; the present invention achieves high accuracy in insulation abnormality judgment, stability in trend identification and adaptability in decision-making strategies by constructing a voltage disturbance shielding mechanism, trend derivative identification and delayed confirmation strategy, and a dual-path safety status decision process with dynamic switching capability, thereby improving the robustness and application breadth of the detection system.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic equipment operation safety monitoring technology, and more specifically, to an intelligent detection system and method for the insulation resistance of a photovoltaic inverter. Background Art

[0002] As the core energy conversion device in photovoltaic power generation systems, the operational safety of photovoltaic inverters is closely related to their electrical insulation performance. To ensure the inverter's long-term stable operation and prevent leakage, breakdown, or equipment damage caused by insulation degradation, it is often necessary to monitor the insulation resistance to ground in real time during operation and issue alarms or protective controls based on the test results.

[0003] In the prior art, the method for determining the insulation resistance of the inverter mostly adopts a single-cycle sampling and fixed threshold comparison mode, that is, the currently detected insulation resistance value is compared with the preset safety limit, and if it is lower than the threshold, it is determined to be an abnormal state. However, this method has the following problems: during the startup or shutdown of the inverter, the bus voltage undergoes a short-term jump due to the capacitor charging and discharging effect, which can easily cause instantaneous distortion of the insulation resistance value. If there is no isolation mechanism, it is easy to trigger a misjudgment. Secondly, the inverter usually operates in outdoor scenes with frequent environmental fluctuations. The ambient temperature and humidity have a significant impact on the insulation characteristics. Their changes can cause periodic disturbances in the resistance measurement value, making the fixed threshold strategy lack stability and adaptability.

[0004] Furthermore, under complex operating conditions such as drastic grid-connected conditions, unstable grid quality, and significant nonlinear load variations, a single-point insulation resistance value cannot effectively identify trending degradation and cannot promptly reflect the evolution of the insulation state. Existing detection methods are particularly prone to failure in the following typical usage scenarios: industrial and commercial distributed power stations where inverters frequently start and stop; tropical and coastal areas with high humidity and temperature or large day-night environmental contrasts; and complex grid-connected environments connected to fluctuating power grids and coexisting inductive loads or remote power feeders. Insulation monitoring methods based on static judgment not only have limited accuracy but also fail to identify potential trends, making it difficult to provide a decision-making basis for hierarchical response control and remote operation and maintenance management. Therefore, an intelligent detection system and method for the insulation resistance of photovoltaic inverters are proposed to address the aforementioned issues. Summary of the Invention

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A photovoltaic inverter insulation resistance intelligent detection method comprises the following steps:

[0007] Obtain the inverter operating status, enter the state transition phase after detecting the start or shutdown event trigger, record the bus voltage change data in real time, and synchronize the state change event with the voltage sampling sequence;

[0008] Through the voltage dynamic convergence modeling method, the corresponding peak point, voltage stable section starting point and convergence time node in the bus voltage curve are extracted, the start and end time intervals of the freezing window are calculated, and the participation of insulation resistance data in the judgment logic within the freezing window is shielded;

[0009] After the freeze window ends, the insulation resistance value is continuously collected, and a derivative calculation is performed on the resistance increments of multiple consecutive sampling cycles. When the resistance value change rate exceeds the set threshold for three or more cycles, it is determined to be an abnormal change trend and the status assessment process is entered;

[0010] The resistance change trend results, ambient temperature, ambient humidity, load power, and grid voltage stability parameters are input into the safety state decision process. The safety state decision process includes two structures: a rule judgment path and a weight fusion path. Path selection is performed during initial configuration or when switching conditions are met. The switching conditions are based on a comprehensive judgment of the change amplitude and stability indicators of multiple input variables. The safety state decision process outputs the operating state level, including normal state, trend warning state, restricted operation state, and emergency shutdown state.

[0011] Corresponding control instructions are executed according to the operating status level. Continuous monitoring is maintained in normal status. Parameter records are reported and remote prompts are issued in trend warning status. Output power is reduced to the set threshold in restricted operation status. In emergency shutdown status, the output circuit is immediately disconnected to ensure electrical insulation safety.

[0012] In a preferred embodiment, the bus voltage change data in the state transition stage obtains a continuous voltage value sequence through a preset sampling frequency, and the voltage dynamic convergence modeling method includes calculating the voltage difference between adjacent sampling points and constructing a voltage change rate curve, locating the peak point and the starting point of the convergence stable interval on the curve, and identifying the convergence time node based on the condition that the voltage change rate is lower than the stability threshold in multiple consecutive sampling cycles; the start and end time intervals of the freezing window are dynamically calculated based on the time distance between the peak point and the convergence time node to automatically set the invalid interval of the insulation resistance detection during the shielding interference period.

[0013] In a preferred embodiment, the resistance change trend result is obtained by performing a derivative calculation on the insulation resistance data within a continuous sampling period. The calculation method is to divide the difference between the resistance value of the current period and the resistance value of the previous period by the length of the sampling period, and the result is the resistance change rate. When the resistance change rate exceeds the set rate threshold for three or more consecutive sampling periods, it is determined to be an abnormal trend state. The rate threshold is determined by the voltage resistance characteristics of the insulating material, the ambient temperature and humidity conditions, and the historical data modeling results, and is set by any of the following methods: a fixed empirical value, a segmented dynamic value, or a logical interval based on fuzzy boundary setting.

[0014] In a preferred embodiment, the calibration behavior of the resistance change trend has a delayed confirmation mechanism. After identifying the initial abnormal trend, the mechanism starts two additional sampling cycles to continue observing the resistance change. If the resistance change rate of the new sampling cycle remains in the same direction and continues to exceed the aforementioned rate threshold, the trend is confirmed to be established. If the rate of any new cycle is lower than the threshold or the change direction is opposite, the calibration behavior is canceled and the system returns to normal.

[0015] In a preferred embodiment, when the safety status decision process executes the weight fusion path, a multivariable weighted scoring function is used to output the status level. The scoring function uses the resistance change trend result, ambient temperature, ambient humidity, load power and grid voltage stability parameters as input variables. Each variable is uniformly normalized and converted into a scoring value before input; the scoring function multiplies each scoring value with a predefined weight factor and sums them to form a status level score, which is distributed between zero and one and is mapped to normal state, trend warning state, restricted operation state or emergency shutdown state; the weight factor is configured through initialization parameters or generated by training of historical operation data.

[0016] In a preferred embodiment, the remote prompt behavior in the trend warning state includes generating a status information package containing five types of data: the current insulation resistance change trend, the average ambient temperature and humidity in the past 30 seconds, the current load power, the grid voltage fluctuation amplitude, and the start and end time of the freeze window, and sending it to the remote monitoring platform through the data upload channel;

[0017] The power reduction behavior under the restricted operating state includes calculating the power reduction target value. The power reduction target value is the product of the current output power and the power reduction factor. The power reduction factor is determined by the maximum change rate in the resistance change trend result and the current ambient temperature and humidity parameters. Its value is less than 1 and has multiple levels. If the grid voltage stability parameter is also lower than the stability threshold, the target power value will introduce a preset voltage attenuation compensation factor to perform an additional reduction.

[0018] In a preferred embodiment, the output circuit interruption control in the emergency shutdown state is implemented through parallel dual-path logic, wherein one path receives the emergency state signal output by the safety state decision process, and the other path receives the auxiliary signal confirmation, including a verification event in which the resistance value is lower than the insulation safety lower limit for three consecutive cycles and the change rate direction is consistent. The disconnection operation is performed only when both paths are activated.

[0019] In a preferred embodiment, the start and end intervals of the freezing time window can be generated using different calculation methods based on a preset strategy, and the strategy includes one of the following three methods: the first method is to collect the voltage change rate curve. When it is detected that the change rate is lower than the voltage stability threshold value in five consecutive sampling periods, the interval is judged to be a voltage convergence segment, the starting point of the freezing window is the peak point of the curve, and the end point is the starting point of the convergence segment, forming a freezing interval between the peak point and the convergence point; the second method is to collect the maximum fluctuation value of the bus voltage in the voltage stability segment. When the fluctuation amplitude does not exceed the set offset threshold, the time interval corresponding to the continuous sampling periods in the collected voltage stability segment is used as the freezing window; the third method is to calculate the average convergence time based on historical operating data, and add a fixed safety margin as the window length, and form a freezing interval starting from the peak point.

[0020] In a preferred embodiment, the safety status decision process includes two structures: a rule determination path and a weight fusion path. The rule determination path outputs the status level based on the conditional combination logic between the input variables; the weight fusion path outputs the status level by weighted calculation using normalized scores and weight factors; the two paths have a dynamic switching mechanism, and the switching condition is that when the change amplitudes of at least two types of input variables exceed their respective stability thresholds at the same time within a continuous observation period, or when the variances of each variable exceed the set composite standard, the path automatically switches from the rule determination path to the weight fusion path.

[0021] In a preferred embodiment, a photovoltaic inverter insulation resistance intelligent detection system includes a state recognition module, an electrical modeling module, a trend assessment module, a safety decision module, and a remote interaction module. These modules work together to construct an intelligent identification and protection response process for insulation performance, specifically including:

[0022] The state recognition module is used to obtain the inverter operating status and bus voltage data in real time, and identify the state transition period corresponding to the startup and shutdown events;

[0023] The electrical modeling module is used to perform convergence modeling on the bus voltage variation curve, extract the peak point and the starting point of the stable section, generate a frozen time window to shield the insulation resistance data within the period, and calculate the resistance change rate to construct a change trend curve;

[0024] The trend assessment module is used to determine the resistance change rate over multiple consecutive sampling periods. If the threshold conditions are continuously met, it is marked as an abnormal trend and input into the subsequent safety determination process together with the ambient temperature and humidity, load power, and grid stability parameters.

[0025] The safety decision module includes two decision structures: rule-based decision paths and weighted fusion paths. It also has a path switching mechanism that selects a decision path based on the fluctuation level of input variables and outputs four operating status levels: normal, trend warning, restricted operation, and emergency shutdown.

[0026] The remote interaction module is used to generate a status data packet containing trend results, environmental information, electrical parameters and frozen window time period in the trend warning state or a higher level state, and send it to the remote monitoring platform to realize remote abnormality perception and system operation and maintenance linkage.

[0027] The technical effects and advantages of the present invention are as follows:

[0028] This invention effectively addresses the problem of detection errors caused by voltage fluctuations during inverter startup or shutdown by constructing a freezing window mechanism based on dynamic bus voltage convergence modeling. By dynamically calculating the freezing window start and end times based on the time interval between the peak point and the convergence time node, insulation resistance data is excluded from the judgment logic during the disturbance period, ensuring that detection is performed only during the voltage stabilization phase. This improves the accuracy of insulation anomaly identification, significantly reduces the probability of misjudgment caused by capacitor charging and discharging and transient disturbances, and enhances the system's anti-interference capability and adaptability.

[0029] The present invention achieves sensitive identification and stable calibration of insulation performance degradation trends through a resistance change trend derivative judgment mechanism and a delayed confirmation strategy. Trend judgment is triggered only when the derivative value exceeds the rate threshold for multiple consecutive sampling cycles. A two-cycle delayed confirmation is introduced to ensure the continuity and directional consistency of trend changes and eliminate false positive judgments caused by short-term fluctuations. This trend judgment logic enables the detection process to have time series recognition capabilities, making it suitable for complex operating conditions where chronic insulation aging and sudden anomalies coexist. It facilitates early intervention in risk control and improves the foresight of system safety level judgments.

[0030] The present invention constructs a dual-structure safety state decision-making process consisting of a rule-based judgment path and a weighted fusion path, and is equipped with a dynamic switching mechanism based on multi-variable fluctuation indicators, so that the state identification strategy has flexible adaptability. Under conditions where variables are operating smoothly, the rule path with high computational efficiency is given priority to ensure response speed; when multiple variables are violently disturbed and the risk of rule failure increases, the system automatically switches to the fusion path, relying on normalized scoring and weighting functions to achieve refined judgment. This dual-path structure improves the accuracy and robustness of judgment while ensuring real-time performance, and can be widely adapted to the insulation safety management needs in different photovoltaic scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings;

[0032] Figure 1 The schematic diagram of the intelligent detection method for insulation resistance of a photovoltaic inverter in the present invention is shown.

[0033] Figure 2 The schematic diagram of the intelligent detection system for insulation resistance of a photovoltaic inverter in the present invention is shown in FIG. DETAILED DESCRIPTION

[0034] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0035] Reference Figure 1 - Figure 2 The following examples were obtained:

[0036] Example 1: The present invention provides an intelligent detection method for the insulation resistance of a photovoltaic inverter. First, a detection pre-mechanism based on operating status perception and voltage dynamic feature recognition is established. During the state transition stage of inverter startup or shutdown, a continuous voltage value sequence is collected and dynamic convergence modeling is performed to extract the peak point and the starting point of the voltage stable section, calculate the start and end time intervals of the frozen window, and shield the influence of the resistance data on the subsequent judgment logic within the window, thereby effectively avoiding misjudgment caused by bus surge voltage changes.

[0037] After entering the effective detection phase, the method continuously samples and calculates the derivative of insulation resistance data to construct resistance trend data. A delayed confirmation mechanism is implemented to identify persistent abnormal trends. This judgment result, along with multiple parameters such as ambient temperature and humidity, load power, and grid voltage stability, is fed into the safety state decision process. The safety state decision process comprises a rule-based decision path and a weighted fusion path, with dynamic path switching capabilities based on the fluctuation amplitude and stability indicators of the input variables. Based on the operating status level output by the decision process, the operating state is categorized as normal, trend warning, restricted operation, and emergency shutdown.

[0038] Each operating state corresponds to a specific control behavior. In the trend warning state, a data packet containing five types of information will be generated and uploaded to the remote monitoring platform. The restricted operating state is controlled by calculating the power reduction target value. The emergency shutdown state uses dual-path verification logic to control the output circuit interruption. In addition to the standard calculation method, the start and end time of the freeze window also supports modeling based on voltage fluctuation amplitude or historical convergence time to achieve multi-strategy adaptation capabilities and improve the robustness and stability of the detection system in complex operating scenarios. Specifically, it includes the following steps:

[0039] The inverter's operating status is acquired, and after detecting a startup or shutdown event trigger, the state transition phase is entered. Bus voltage change data is recorded in real time, and the state change event is time-synchronized and bound to the voltage sampling sequence. A complete operating status perception framework is constructed to enable the detection behavior to have the ability to identify electrical state switching. By capturing state switching events such as startup and shutdown, and combining them with continuous sampling data of the bus voltage, the system can synchronously match event nodes with voltage fluctuation behavior, thereby providing an accurate reference on the timeline for the subsequent identification of voltage disturbance characteristics. This time synchronization mechanism forms the triggering basis of the detection process, giving the entire detection strategy dynamic response capabilities and the ability to be tightly coupled with the inverter's operating cycle to avoid erroneous startup of the judgment logic during the state transition phase.

[0040] Through voltage dynamic convergence modeling, the corresponding peak points, starting points of the voltage stable section, and convergence time nodes in the bus voltage curve are extracted, the start and end time intervals of the freeze window are calculated, and the participation of insulation resistance data in the judgment logic within the freeze window is shielded. A modeling mechanism for voltage disturbances is introduced to analytically identify the non-steady-state characteristics of the voltage during state changes and determine the effective coverage interval of the interference cycle. By extracting peak points and convergence sections, the system can accurately identify the critical process of the bus voltage transition from violent fluctuations to stability, and construct a freeze window to temporarily exclude the insulation resistance test results in this stage, preventing abnormal resistance value disturbances caused by physical phenomena such as bus capacitor charging and discharging from interfering with the judgment results. This mechanism provides a "data isolation period" for the judgment logic, ensuring that detection behavior is carried out under the premise of controllable signal quality, enhancing the detection system's immunity to electrical disturbances.

[0041] After the freeze window ends, the insulation resistance value is continuously collected, and a derivative calculation is performed on the resistance increment of multiple consecutive sampling cycles. When the resistance value change rate continues to exceed the set threshold for three or more cycles, it is determined to be an abnormal change trend and enter the status assessment process; trend indicators are extracted from the dynamic change behavior of the insulation resistance to replace the traditional static threshold judgment logic, thereby achieving early identification of potential degradation processes. By constructing a resistance change rate indicator through derivative calculation, the system can determine whether the insulation performance is steady, slowly changing, or continuously declining. If the resistance value shows a rate exceeding the limit in the same direction for multiple consecutive cycles, it means that the equipment has a structural degradation risk and needs to enter the next stage of the comprehensive assessment process in a timely manner. This strategy emphasizes trend recognition capabilities, improves the system's sensitivity to the insulation performance degradation process, and avoids single-point abnormality misjudgment and delayed response.

[0042] Resistance trend results, ambient temperature, humidity, load power, and grid voltage stability parameters are input into the safety state decision process. The process consists of two structures: a rule-based decision path and a weighted fusion path. Path selection is performed during initial configuration or when switching conditions are met. Switching conditions are based on a comprehensive assessment of the magnitude of changes in multiple input variables and stability indicators. The safety state decision process outputs operating status levels, including normal, trend warning, restricted operation, and emergency shutdown. A multi-source fusion, path-switching intelligent judgment system is constructed to perform a multi-dimensional comprehensive assessment of insulation status. Electrical trend information is combined with key variables such as environmental parameters, load information, and grid characteristics. The current operating status is classified using predefined judgment logic and a weighted fusion model. The rule-based decision path addresses simple scenarios with clear rules, while the weighted fusion path handles dynamic conditions with complex variable interactions. Path switching can be implemented based on the degree of input fluctuation, ensuring that the judgment strategy maintains accuracy and stability across various operating environments. This mechanism enhances the intelligence and adaptability of state judgment and is the core of implementing dynamic protection strategies.

[0043] Corresponding control instructions are executed according to the operating status level. Continuous monitoring is maintained in normal conditions. Parameter records are reported and remote prompts are issued in trend warning conditions. Output power is reduced to the set threshold in restricted operating conditions. In emergency shutdown conditions, the output circuit is immediately disconnected to ensure electrical insulation safety. The status identification results are linked with the actual operating control logic to establish a closed-loop control mechanism from identification to response. In normal conditions, the system maintains routine monitoring without intervention. In trend warning conditions, the system generates data packets and transmits them to the remote monitoring platform, enabling remote alarms and trend retention. In restricted operating conditions, the system proactively reduces output based on the calculated power target value to avoid further degradation caused by excessive electrical pressure. In emergency shutdown conditions, the system uses dual-path judgment logic to ensure the accuracy of interruption behavior, promptly cutting off the electrical output path to ensure the safety of personnel and equipment. This step accurately maps the judgment results into graded response behaviors and is a key link in achieving risk closed-loop control and adaptive operating strategies.

[0044] Intelligent testing of PV inverter insulation resistance begins with determining the inverter's operating status. This involves collecting and identifying information such as the inverter's control instructions, output behavior, and status registers to determine whether the inverter is currently in the "startup," "operational," or "shutdown" state. When a startup or shutdown event is detected—that is, when the operating state changes from non-operating to operating or vice versa—the system automatically enters the state transition phase. This state transition phase, characterized by strong disturbances and high instability, is when electrical quantities such as busbar voltage and current experience dramatic fluctuations due to interruptions. This phase provides a critical window for determining the system's dynamic behavior.

[0045] During the state transition phase, the system records bus voltage changes in real time. Bus voltage is the voltage of the main power path connecting the PV panels to the inverter, reflecting the current electrical status of the equipment. Voltage data is collected by obtaining a continuous sequence of voltage values at a preset sampling frequency. This involves setting a fixed sampling period, such as recording the bus voltage every 5 milliseconds, to form a raw voltage sequence with continuous time and clear amplitude changes.

[0046] The system then applies a voltage dynamic convergence modeling approach to model and analyze the voltage data. This "modeling approach" involves structurally extracting the temporal trend of voltage changes through digital processing. This process involves calculating the voltage difference between adjacent sampling points—that is, performing a differential operation on two consecutive voltage sampling points to determine the instantaneous amplitude of the voltage change. This voltage difference is then used to construct a voltage change rate curve, which is used to analyze the severity and trend of voltage fluctuations.

[0047] On this curve, the system further locates the peak point, which is the highest or lowest point in the voltage change process, typically corresponding to the extreme voltage state at the beginning or end of a surge. It also locates the starting point of the convergence stable interval, which is the first time point when the voltage begins to stabilize. The basis for determining these two points is that when the voltage change rate falls below the stability threshold for multiple consecutive sampling periods, that is, the voltage change rate is less than a preset rate of change (such as no more than five volts per second), the system considers that the segment has entered the steady-state interval and thus identifies the convergence time point.

[0048] With these two key time points—the peak and convergence points—the system calculates the start and end time intervals of the freeze window, which are the intervals between the two points. This time interval, known as the freeze window, automatically sets the insulation resistance test invalidation interval during interference suppression. In other words, within this freeze window, the system ignores or blocks insulation resistance test results to prevent misjudgment of electrical offsets due to severe bus voltage disturbances.

[0049] For example, during an inverter startup, the system samples the bus voltage every 5 milliseconds, observing a rapid rise to a peak before gradually decreasing and entering a stable phase. The peak occurs 45 milliseconds after startup, and the convergence occurs at 190 milliseconds. The freeze window is set from 45 to 190 milliseconds after startup. During this period, any changes in insulation resistance are not used as a basis for status determination. Only test results outside the freeze window are used by the system to determine insulation trends, ensuring the accuracy of the judgment logic and its ability to resist interference.

[0050] This mechanism is connected with the subsequent resistance derivative calculation, resistance trend judgment, and safety status decision-making process, providing a reliable early anti-disturbance strategy for the entire detection method. It also forms a closed-loop logic chain with subsequent responses such as power reduction and disconnection control, and is a basic step to achieve accurate detection and stable operation.

[0051] The resistance trend result is calculated by performing a derivative calculation on the insulation resistance data within a continuous sampling period. Insulation resistance data refers to the inverter-to-ground insulation resistance value collected by the system during each fixed sampling period, typically in megohms. A continuous sampling period refers to multiple consecutive time periods collected by the system at a set sampling frequency, such as five consecutive sampling periods with a 10 millisecond period.

[0052] The derivative calculation method is to divide the difference between the resistance value in the current cycle and the resistance value in the previous cycle by the sampling period length. Its physical meaning is the rate of change of the resistance value per unit time, also known as the "resistance change rate." This rate value is a scalar, typically expressed in megaohms per second. Positive values indicate improved insulation, while negative values indicate decreased insulation. Larger absolute values indicate more dramatic changes. This calculation method is a first-order derivative approximation and is suitable for digital control processing in industrial field environments.

[0053] The system uses the resistance change rate for three or more consecutive sampling periods as an analysis window. If the rate value exceeds the set rate threshold within this window, the system determines that the abnormal trend is in place. "Three or more consecutive periods" means that the abnormal rate behavior remains stable over the three sampling periods, eliminating misjudgments caused by short-term random fluctuations. The "set rate threshold" is the numerical threshold for determining the existence of an abnormal trend. The rate must continuously exceed this value to be considered a trend change.

[0054] This rate threshold is determined by a combination of factors, including:

[0055] Voltage withstand characteristics of insulating materials: Different materials have different physical responses to insulation degradation under high voltage, and the voltage drop rate is used as a reference for the lower threshold;

[0056] Ambient temperature and humidity conditions: In high-humidity and high-temperature environments, the insulation value drops faster, so a dynamic threshold needs to be set to adapt to environmental disturbances;

[0057] Historical data modeling results: After running for a period of time, the system can generate a statistical distribution diagram of the resistance change rate, and then fit a reasonable threshold range.

[0058] To adapt to various working conditions, the threshold can be selected from the following three methods:

[0059] Fixed empirical value: Based on experiments or industry experience, a fixed rate threshold is set, such as a drop of no more than two megaohms per second;

[0060] Segmented dynamic value: Divide segments by time, environment or device type, and use different thresholds for each segment;

[0061] Logical interval based on fuzzy boundary setting: Set a slowly changing interval and use a fuzzy judgment strategy between logical judgments of "highly credible trends" and "suspicious trends" to improve the algorithm's fault tolerance and judgment sensitivity.

[0062] For example, in a certain operating environment, with a sampling period set at 100 milliseconds, the system detected resistance values of 100 megohms, 90 megohms, and 75 megohms three times within 300 milliseconds. The calculated value was a decrease of 50 megohms per second, far exceeding the empirical rate threshold of 20 megohms per second. This triggered an abnormal trend assessment and initiated the subsequent status assessment process. This approach not only improves the foresight of detection but also strengthens the ability to continuously perceive the insulation degradation process, forming a key foundation for implementing intelligent judgment strategies and multi-level safety response.

[0063] In the intelligent detection method for photovoltaic inverter insulation resistance presented herein, to improve the stability and accuracy of trend determination, the system implements a resistance trend calibration behavior after generating resistance trend results. Calibration behavior involves the system identifying a trend state when a set of judgment conditions are met, and this serves as input for subsequent state assessment processes. To prevent misjudgments caused by transient disturbances or occasional jumps, after the initial criteria for an abnormal trend are met, the system does not immediately calibrate the trend as valid. Instead, it performs secondary verification through a delayed confirmation mechanism.

[0064] The delayed confirmation mechanism is a strategy structure that is automatically triggered after a preliminary abnormal trend is identified in the previous phase. It operates by initiating two additional sampling cycles. This extends two additional sampling cycles beyond the three consecutive sampling cycles that met the initial judgment criteria, continuing to observe subsequent insulation resistance changes. These two sampling cycles extend consecutively in time, typically using the same sampling period as the previous one (for example, a 100 millisecond period per cycle, resulting in a 200 millisecond delayed confirmation period).

[0065] During this delay phase, the system calculates the resistance change rate within each new sampling period in real time. The definition is the same as above, that is, the difference between the resistance value of the current period and the resistance value of the previous period divided by the length of the sampling period. The result is the insulation resistance change rate per unit time, in megaohms per second.

[0066] The delayed confirmation mechanism sets two criteria:

[0067] The first is directional consistency: the rate of change of resistance over the two new cycles should be consistent with the previously determined trend direction. The trend direction refers to whether the resistance value is continuously decreasing or increasing. For example, if the resistance value has been decreasing over the previous three cycles, it should also be decreasing over the two new cycles.

[0068] The second requirement is threshold persistence: the resistance change rate must continue to exceed the aforementioned rate threshold for both new cycles. The rate threshold is set as previously disclosed and is typically calculated based on the insulation material, electrical environment, and historical models.

[0069] If the resistance change rate in the new sampling cycle remains in the same direction and continues to exceed the aforementioned rate threshold, the system deems the resistance change not an occasional disturbance but a continuous trend, confirms the trend, and inputs it as the official trend result into the safety state decision process. Conversely, if the rate in any new cycle falls below the threshold or changes in the opposite direction, indicating a trend interruption or reversal, the system deems the previous abnormal trend to be a false trigger, cancels the calibration action, and returns to normal.

[0070] For example, during a specific detection process, the system sampling period is set to every 100 milliseconds. The resistance values in the first three cycles are 100 megohms, 80 megohms, and 65 megohms, respectively. The corresponding rates are all decreasing and exceed the set threshold of 20 megohms per second, triggering a preliminary abnormal trend judgment. At this point, the system enters the delayed confirmation phase. In the next two sampling cycles, the resistance values are 50 megohms and 35 megohms, respectively. If the rate continues to be negative and the absolute value exceeds the threshold, the system confirms that the trend has been established. If the resistance value rises to 70 megohms or the rate is less than 10 megohms per second during one of the cycles, the trend is deemed to have failed and the system returns to a normal state. The core significance of this mechanism is to reduce false alarms and enhance robustness. In practical applications, electrical systems are susceptible to transient interference, environmental fluctuations, and load switching. Without a delayed confirmation mechanism, occasional behavior can be misinterpreted as a trend, resulting in frequent jumps in control response. By introducing a delayed period for judgment, the system bases trend identification on sustained behavior, thereby improving the stability, accuracy, and interference resistance of state recognition.

[0071] The safety state determination results are used to drive different control strategies. The determination process consists of two structures: a rule-based determination path and a weighted fusion path. When the safety state decision process executes the weighted fusion path, the system no longer determines the state level based on pre-set rule logic. Instead, it constructs a function model based on multiple parameter inputs and calculates the state using continuous numerical values.

[0072] In this path, the system uses a multivariate weighted scoring function to output the status level. The scoring function is to weight multiple input factors according to their importance and output a unified scoring value, called the status level score, to reflect the comprehensive operating safety level of the current equipment.

[0073] The scoring function uses resistance change trend results, ambient temperature, ambient humidity, load power and grid voltage stability parameters as input variables. Among them:

[0074] Resistance change trend result: refers to the resistance change rate index calculated by the aforementioned derivative, which represents the speed and direction of insulation performance change;

[0075] Ambient temperature and ambient humidity: These represent the real-time temperature and humidity of the inverter's operating environment, which have a significant impact on insulation degradation.

[0076] Load power: refers to the real-time power value of the power load driven by the current inverter, usually obtained by multiplying the voltage and current;

[0077] Grid voltage stability parameter: Indicates the fluctuation amplitude or fluctuation rate of the grid voltage over a period of time, and is used to assess the degree of disturbance in the external electrical environment.

[0078] All of the above variables are uniformly normalized before input and converted into score values. Normalization converts different physical quantities (such as temperature, power, and voltage fluctuation) into dimensionless proportional values between zero and one to facilitate weighting within the function. Typical methods used include maximum and minimum value normalization, sliding window normalization, or empirical range normalization.

[0079] The scoring function multiplies each score by a predefined weighting factor and then sums them to form a state grade score. The score is a normalized result, while the weighting factor is a coefficient set in advance or adaptively learned, which represents the influence of each input variable on the state grade. For example, if the resistance trend has a greater impact than the ambient humidity, the resistance trend will receive a higher weight.

[0080] This score ranges from 0 to 1 and is mapped to normal status, trend warning status, restricted operation status, or emergency shutdown status. The specific mapping relationship can be set according to segmented rules, such as: 0 to 0.25 is normal status; 0.251 to 0.5 is trend warning status; 0.51 to 0.75 is restricted operation status; and above 0.75 is emergency shutdown status. This interval division can be adjusted according to the actual operational safety tolerance.

[0081] The weighting factors are generated through initialization parameter configuration or training from historical operating data. This setting method is used to determine the influence ratio of each input variable in the state level scoring function and is the core parameter for constructing the mathematical model of the scoring function. First, initialization parameter configuration means that in the early stage of system design or deployment, a set of fixed weight values is obtained in advance through engineering experience, industry standards or simulation experiments, and written into the system's initial configuration file. This method is suitable for equipment scenarios where operating data has not yet been accumulated. For example, for initially installed inverter equipment, the initial weighting factors of each variable can be set according to the following rules: resistance change trend result: 0.45; ambient temperature: 0.15; ambient humidity: 0.1; load power: 0.15; grid voltage stability parameter: 0.15; such initial configuration values can be set differently according to the equipment type, installation location or operation and maintenance preferences, and can also be pre-stored in the control chip or set through the user interface parameter menu to ensure the executable nature of the scoring function.

[0082] Historical operation data training generation refers to recording the historical correspondence between input variables and status level outputs during the long-term operation of the system, and automatically optimizing the distribution ratio of weight factors through training algorithms to make the scoring function more consistent with actual operation characteristics and safety judgment goals. The specific training method may include the following steps:

[0083] Data collection: In each detection cycle, the system automatically saves the resistance change trend results, ambient temperature, ambient humidity, load power, grid voltage stability parameters and the operating status level finally determined by the system at that time;

[0084] Sample construction: Use the five input variables as feature vectors and the state level as labels to construct a historical training sample set;

[0085] Parameter fitting: Use multivariate regression, gradient descent, or other machine learning optimization algorithms to fit weight factors so that the output of the scoring function is closest to the historical label status level;

[0086] Result verification and application: The accuracy of the training results is evaluated through cross-validation and other methods. Under the premise of meeting the judgment accuracy requirements, the weight parameters are written into the current model to replace the initialization weight factors as the effective scoring basis in the current operation cycle. For example, after thirty days of operation data learning, the system found that the grid voltage stability was more sensitive to state abnormality prediction, and automatically adjusted its weight from the initial 0.15 to 0.25, while moderately reducing the weight of the ambient humidity. This behavior does not require manual intervention, and the system can train and automatically switch in the background to achieve online evolution of the scoring model. Through the combined mechanism of initialization configuration and training generation, the weight factor setting method in the present invention not only meets the availability of rapid deployment of the system in the initial stage, but also ensures the accuracy and adaptability of the model in long-term operation, thereby enhancing the generalization performance, stability and prediction accuracy of the state level evaluation function.

[0087] For example, in a certain field application, the resistance change trend score was 0.85, the ambient temperature score was 0.6, the humidity score was 0.7, the load power score was 0.5, and the voltage stability score was 0.95. The corresponding weight factors were 0.35, 0.1, 0.1, 0.2, and 0.25, respectively. The final state grade score was the sum of the products of the above items, which was 0.78, falling into the "emergency shutdown state" range, and the system immediately entered the disconnect loop response. This method has significant numerical adaptability, model trainability, and continuity in state judgment. It avoids the problem of rule paths failing under multivariable coupling conditions and provides dual guarantees of accuracy and stability for comprehensive judgment of dynamic changes in insulation performance.

[0088] The status level determination triggers corresponding control actions to ensure safe equipment operation and provide maintenance support. After identifying different operating status levels, the system will implement remote notification actions in trend warning states and power reduction actions in restricted operating states, corresponding to the intelligent response strategies for the two abnormal status levels of "trend warning state" and "restricted operating state."

[0089] Remote notification behavior in the trend warning state means that when the status level is judged to be a trend warning state, the system automatically generates a set of data reflecting the current electrical status change trend and environmental change parameters for remote monitoring and analysis. The core steps of this behavior include: generating a status information package containing five types of data: the current insulation resistance change trend, the average ambient temperature and humidity over the past 30 seconds, the current load power, the grid voltage fluctuation amplitude, and the start and end time of the freeze window. Among them:

[0090] Current insulation resistance change trend: the most recent trend identification result, including the change direction (increase or decrease) and change rate;

[0091] Average ambient temperature and humidity over the past 30 seconds: The system collects all ambient temperature and humidity samples over the past 30 seconds and calculates their arithmetic average to smooth out short-term disturbances.

[0092] Current load power: the inverter output power value in the current detection cycle;

[0093] Grid voltage fluctuation amplitude: the difference between the maximum and minimum grid voltage within the set time window, reflecting the grid power supply stability;

[0094] Freeze window start and end time: The freeze detection mask period corresponding to the startup or shutdown event is used to support remote state reconstruction.

[0095] The five types of data described above are packaged into status information packets, which are then sent to the remote monitoring platform via a data upload channel. This data upload channel can be a local area network, wired communication link, wireless cellular network, or other data transmission method. The remote monitoring platform is a backend system with data reception, analysis, visualization, and alarm management capabilities. It is used to remotely identify equipment operating trends, perform predictive maintenance, and construct status evolution curves. By triggering remote notifications and transmitting data, operations and maintenance personnel can proactively intervene before a fault occurs.

[0096] When the status level is judged to be a restricted operating state, the system will perform power reduction to reduce the inverter's electrical output load, alleviate the trend of insulation degradation, and extend the equipment's safe operation time. This behavior includes: calculating the power reduction target value, which is the product of the current output power and the power reduction factor. Among them:

[0097] Current output power: The real-time output power of the device in the current cycle, in watts or kilowatts;

[0098] Derating factor: A coefficient less than one used to calculate the target power, indicating the ratio limit of the allowed output power.

[0099] The power reduction factor is determined by the maximum rate of change in the resistance change trend result and the current ambient temperature and humidity parameters. In other words, this factor is not a fixed setting, but is determined by the electrical parameters and environmental factors:

[0100] The maximum change rate in the resistance change trend result refers to the maximum value of the resistance drop rate in the most recent cycles, which represents the severity of insulation degradation;

[0101] Current ambient temperature and humidity parameters: that is, the temperature and humidity values within the current detection period, which have a nonlinear effect on insulation performance.

[0102] Based on the two input variables mentioned above, the system can set multiple power reduction levels, which can be set based on fuzzy reasoning or other preset logic. For example: if the resistance drop rate is low and the temperature and humidity are normal, the factor is 0.9; if the resistance drops quickly and the ambient humidity is high, the factor is reduced to 0.75; if the temperature and humidity are extreme at the same time, the factor can be reduced to 0.5.

[0103] If both grid voltage stability parameters fall below the stability threshold, meaning the grid voltage fluctuation exceeds the equipment's safety limit, the system applies an additional voltage reduction based on the target power using a pre-set voltage attenuation compensation factor. This factor is a predefined coefficient, typically set below 0.9, to further reduce load pressure when grid disturbances and internal equipment failure risks combine.

[0104] For example, if the current output power is 5 kilowatts and the resistance change rate is high, the power reduction factor is set to 0.6, and the target power is 3 kilowatts. At this time, the grid voltage stability is poor, and the compensation factor is 0.9, so the final output power is 3,000 times 0.9, or 2,700 watts. This power reduction behavior allows the system to proactively reduce system output when it detects that insulation risk has not yet reached disconnection requirements but is showing signs of developing. This maintains power generation continuity while controlling risk boundaries, and is a key strategy for transitioning from "fault-prone operation" to "active preemptive control."

[0105] When the system identifies that the equipment is in an emergency shutdown state, in order to ensure the safety of electrical insulation, the electrical output path must be immediately interrupted to prevent breakdown, leakage or personal safety risks caused by further operation. To this end, a disconnection mechanism with safety redundancy is constructed, that is, the output circuit interruption control is implemented through parallel dual-path logic. Output circuit interruption control refers to controlling the conduction state between the AC output part of the inverter and the external load, so that it cuts off the output under specific conditions to achieve electrical isolation. This action is generally completed by the control unit driving the relay, contactor or semiconductor switching device. The disconnection operation in the present invention does not rely on a single judgment result, but is implemented through parallel dual-path logic, that is, the disconnection action must simultaneously meet the activation conditions of the two judgment paths to be executed, thereby building logical redundancy to avoid false triggering.

[0106] One of the paths receives the emergency state signal output by the safety state decision process. This signal is generated by the aforementioned weight fusion path or rule judgment path after determining that the system is in an "emergency shutdown state", indicating that the system has completed all input variable analysis and determined that there is an unacceptable insulation risk in the current state, and hardware disconnection must be performed immediately. This signal is the first basis for judgment and has the decision-making power of global state information. The other path receives auxiliary signal confirmation. This auxiliary path is independent of the main path judgment logic and is dedicated to "bottom-level verification" of key physical quantities to further ensure that the actual detection results have indeed reached the irrecoverable threshold, thereby preventing unnecessary disconnection operations caused by misjudgment of the main path. The source of the auxiliary signal includes a composite judgment condition:

[0107] A verification event occurs when the resistance value is below the insulation safety lower limit for three consecutive cycles, and the rate of change is consistent. This condition indicates that the insulation resistance values collected by the system during three consecutive sampling cycles are all below a preset minimum allowable resistance value. For example, the safety lower limit can be set to 20 megohms. A consistent rate of change indicates that the resistance value trend is downward over the three cycles, with no reversals, meaning the derivative values are in the same direction. This compound condition is designed to determine whether the current insulation condition is deteriorating and the value has fallen below the safety limit. It logically binds the "trend" and "absolute value" indicators.

[0108] The disconnection operation is only executed when both paths are activated. In other words, the system will only activate the loop interruption control logic when the status level judgment path issues an "emergency stop" signal and the auxiliary judgment path confirms that the actual detection parameters meet the dual characteristics of deterioration and exceeding the limit. Otherwise, even if the main path outputs an emergency signal, if the auxiliary judgment conditions are not met, the disconnection operation will not be executed temporarily, and vice versa.

[0109] For example, the system sampling period is every 100 milliseconds, and the current output power is 4 kilowatts. During the most recent three sampling periods, the resistance values were 18 megohms, 15 megohms, and 12 megohms, respectively—all below the safety lower limit of 20 megohms. The derivative values were all negative, indicating a continuous decrease, satisfying the auxiliary signal condition. Simultaneously, the safety state decision process, calculated using a scoring function, outputs the state level as "emergency shutdown." At this point, both channels are activated, the disconnection logic takes effect, the relay operates, and the output channel is immediately disconnected. This dual-channel structure ensures a rapid disconnect response in true emergencies while reducing the risk of false disconnections. This is especially true in the case of short-term fluctuations, environmental interference, and other non-persistent risks. The system avoids unnecessary interruptions, improving the accuracy of protection control and system stability.

[0110] To effectively shield the inverter from transient interference caused by sharp bus voltage fluctuations during startup and shutdown, the system needs to establish a freeze time window to temporarily block the use of insulation resistance data during this period in determining the inverter's status. To adapt to different electrical conditions and operating scenarios, this freeze time window is not set to a fixed value but rather features a strategic, dynamic configuration mechanism.

[0111] Specifically, the start and end intervals of the freeze time window can be generated using different calculation methods based on preset strategies. This means that the system can dynamically construct the freeze window boundaries by selecting different strategy types through operating parameter settings. The start and end intervals here refer to the start and end times of the freeze window. The window is essentially a time period during which resistance data is excluded.

[0112] The strategy includes one of the following three methods. That is, the system only uses one of the three methods to build the window based on the device configuration or operating conditions. The three methods are:

[0113] The first method involves collecting a voltage rate of change curve. This involves high-frequency sampling of the bus voltage and calculating the voltage difference between each sampling point and the adjacent sampling points to form a voltage rate of change curve. The voltage rate of change refers to the magnitude of the voltage change per unit time, typically measured in volts per second. Within this curve, the system determines whether the rate of change falls below the voltage stability threshold for five consecutive sampling periods. The stability threshold is a pre-set rate threshold representing the maximum acceptable voltage fluctuation rate, such as 10 volts per second. If the rate of change falls below this threshold for five consecutive periods, the system determines that this section is a voltage convergence period, indicating that the voltage has gradually transitioned from an unstable state to a stable state. The system then sets the freeze window starting point to the peak of the curve—the time point at which the maximum or minimum value occurs during the voltage fluctuation process—and the end point to the starting point of the convergence period—the first time point of the continuous stable section. This freeze window forms the interval between the peak and the convergence point. This method emphasizes the characteristics of the voltage fluctuation curve and accurately covers the disturbance period.

[0114] The second method collects consecutive sampling cycles within the voltage stability segment and calculates the maximum bus voltage fluctuation within that period. If the fluctuation amplitude does not exceed a set excursion threshold, the time interval corresponding to these consecutive sampling cycles is defined as a freeze window. The bus voltage stability segment refers to the voltage period after the system enters stable operation, during which the sampled voltage fluctuations are minimal. The system analyzes the difference between the maximum and minimum voltage values within this segment and defines it as the voltage fluctuation amplitude. If this amplitude does not exceed the set excursion threshold, meaning the voltage fluctuation is within a safe tolerance (for example, less than five volts), the system deems the voltage within this segment stable. In this case, the system uses the time interval corresponding to the consecutive sampling cycles within the voltage stability segment as the freeze window, providing a safety buffer after the disturbance ends. This method has more relaxed judgment criteria than the first method and is suitable for scenarios where the fluctuation amplitude of the equipment operating point is minimal. The excursion threshold is factory-configured or user-defined to determine whether the fluctuation is acceptable.

[0115] The third method is to calculate the average convergence time based on historical operating data and add a fixed safety margin as the window length. Historical operating data refers to the voltage fluctuation process data recorded after the equipment has experienced multiple startup, shutdown and other events in its past operation. The system performs statistical analysis on this data and extracts the average time required from the voltage peak to the stable point each time, which is recorded as the average convergence time. To ensure safe coverage, the system adds a predefined fixed safety margin, such as twenty milliseconds, to strengthen the shielding boundary. The final freezing window starts from the peak point and continues the length of "average convergence time + margin" to form a freezing interval. This method is suitable for equipment trained with big data, has on-site self-adaptation capabilities, and can be used to deploy common models for similar equipment.

[0116] For example, in a certain field environment, the bus voltage undergoes a dramatic change after startup, with the voltage peak occurring at the 60th millisecond. Thereafter, the rate of change is less than 8 volts per second for five consecutive sampling cycles (each cycle is 10 milliseconds), meeting the conditions of the first method. The system sets the 60th millisecond as the starting point and the 110th millisecond as the end point, with a freeze window length of 50 milliseconds. The insulation resistance data collected within this window does not participate in subsequent state level judgments. By providing the above three strategies and implementing a mutually exclusive selection structure, the system has the ability to flexibly generate freeze windows, which can be called according to actual scenarios, user configurations, or control logic. This improves the coverage of abnormal disturbances, judgment accuracy, and environmental adaptability of the freezing mechanism without increasing the sampling burden.

[0117] When the system identifies whether there is a downward trend in insulation performance, it needs to conduct parallel analysis of multiple influencing factors to output an accurate and reliable operating status level. To achieve this goal, a judgment framework with a dual-structure path is constructed, that is, the safety state decision process includes two structures: a rule judgment path and a weight fusion path. This structure is flexible in control logic and adaptable in implementation. It is an important component of the present invention to improve the reliability of state identification. The rule judgment path refers to the system outputting the state level based on the conditional combination logic between the input variables, that is, setting a set of empirical or expert rules for logical combination judgment of the specific value relationship between the input parameters. For example: if the resistance change trend result exceeds the warning threshold and the ambient humidity is higher than the set standard, it is judged as a trend warning state; if the resistance rate drops sharply and the grid voltage fluctuates abnormally, a restricted operating state is output. This path is suitable for scenarios where the variable change law is clear and the boundaries can be described in a regular manner. The advantages are simple calculation and rapid response, and it is suitable for deployment on low-computing power terminals.

[0118] The weighted fusion path provides the system with another evaluation capability, outputting the status level through a weighted calculation method using normalized scores and weight factors. The specific process includes: uniformly normalizing all input variables (resistance change trend results, ambient temperature, ambient humidity, load power, and grid voltage stability parameters) and converting them into a score between zero and one; multiplying each score by the corresponding predefined weight factor and weighted summing them to output a status level score; status level scores falling into different intervals are mapped to normal status, trend warning status, restricted operation status, or emergency shutdown status. This path is suitable for scenarios where variable interactions are complex, relationships are nonlinear, or difficult to describe in a regular manner. It can provide more continuous status output results and adapt to complex environments.

[0119] In order to improve the overall adaptability and judgment accuracy of the method, the system is designed with two paths equipped with a dynamic switching mechanism, that is, the system can select the applicable path during operation according to the fluctuation of the input parameters. The specific switching condition is that the change amplitude of at least two types of input variables exceeds their respective stability thresholds at the same time during the continuous observation period, or the variance of each variable exceeds the set composite standard. Among them: the change amplitude refers to the maximum change of the variable within a period of time; the stability threshold is the maximum allowable fluctuation limit set by the system, such as temperature fluctuation exceeding three degrees Celsius and load power change exceeding twenty percent; the variable variance represents the degree of discreteness of the value of the variable within a certain time window, which measures volatility; the composite standard is set as a set of variance threshold standards set for multiple variables to ensure that the overall judgment is based on multiple dimensions and is controllable. The design of automatically switching from the rule judgment path to the weight fusion path prevents the system from making incorrect state judgments due to rule failure when encountering complex operating environments such as instability, abnormal fluctuations, and rules that are difficult to work.

[0120] While the rule-based path offers fast response and minimal computational effort, it's prone to misjudgment when variables fluctuate erratically or when logical relationships are unclear. The weighted path, while accurate, is computationally complex and places a high system load. Therefore, the dual-path complementary design allows for dynamic allocation of judgment resources. PV inverters are often deployed in areas with variable environments, such as high humidity, high altitudes, and areas with severe grid interference, where multiple parameter disturbances coexist. This mechanism automatically identifies variable fluctuation characteristics and selects a more appropriate state assessment method to improve robustness. When the system is operating stably, the rule-based path is prioritized to ensure rapid response. Once variables fluctuate dramatically, the path automatically switches to the weighted path to ensure the accuracy and consistency of judgment results and avoid frequent state changes. The path switching mechanism can be enabled based on system configuration under different O&M strategies, computing resources, or operating models, without being restricted to a fixed architecture, improving the versatility and engineering practicality of the overall technical solution.

[0121] To illustrate, in one operational scenario, the system identifies that the resistance change trend score has fluctuated from 0.5 to 0.9 over the past five sampling cycles, the ambient temperature has fluctuated by 6 degrees Celsius, and the humidity has varied by 30 percent, exceeding the set stability threshold. The system automatically switches from the rule-based decision path to the weighted fusion path, recalculates the status score within the fusion path, outputs an "emergency shutdown status," and triggers the disconnect logic. If this switch is not made, the rule-based path, not covering this variable combination, might determine the status as a "trend warning status," resulting in a delayed response.

[0122] Example 2: A photovoltaic inverter insulation resistance intelligent detection system includes a state recognition module, an electrical modeling module, a trend assessment module, a safety decision module, and a remote interaction module. These modules collaborate to construct an intelligent insulation performance identification and protection response process, specifically including:

[0123] The state recognition module is used to obtain the inverter operating status and bus voltage data in real time, and identify the state transition period corresponding to the startup and shutdown events;

[0124] The electrical modeling module is used to perform convergence modeling on the bus voltage variation curve, extract the peak point and the starting point of the stable section, generate a frozen time window to shield the insulation resistance data within the period, and calculate the resistance change rate to construct a change trend curve;

[0125] The trend assessment module is used to determine the resistance change rate over multiple consecutive sampling periods. If the threshold conditions are continuously met, it is marked as an abnormal trend and input into the subsequent safety determination process together with the ambient temperature and humidity, load power, and grid stability parameters.

[0126] The safety decision module includes two decision structures: rule-based decision paths and weighted fusion paths. It also has a path switching mechanism that selects a decision path based on the fluctuation level of input variables and outputs four operating status levels: normal, trend warning, restricted operation, and emergency shutdown.

[0127] The remote interaction module is used to generate a status data packet containing trend results, environmental information, electrical parameters and frozen window time period in the trend warning state or a higher level state, and send it to the remote monitoring platform to realize remote abnormality perception and system operation and maintenance linkage.

[0128] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0129] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0130] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0131] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0132] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A photovoltaic inverter insulation resistance intelligent detection method, characterized in that: The following steps are involved: Obtain the inverter operating status, enter the state transition phase after detecting the start or shutdown event trigger, record the bus voltage change data in real time, and synchronize the state change event with the voltage sampling sequence; Through the voltage dynamic convergence modeling method, the corresponding peak point, voltage stable section starting point and convergence time node in the bus voltage curve are extracted, the start and end time intervals of the freezing window are calculated, and the participation of insulation resistance data in the judgment logic within the freezing window is shielded; After the freeze window ends, the insulation resistance value is continuously collected, and a derivative calculation is performed on the resistance increments of multiple consecutive sampling cycles. When the resistance value change rate exceeds the set threshold for three or more cycles, it is determined to be an abnormal change trend and the status assessment process is entered; The resistance change trend results, ambient temperature, ambient humidity, load power, and grid voltage stability parameters are input into the safety state decision process. The safety state decision process includes two structures: a rule judgment path and a weight fusion path. Path selection is performed during initial configuration or when switching conditions are met. The switching conditions are based on a comprehensive judgment of the change amplitude and stability indicators of multiple input variables. The safety state decision process outputs the operating state level, including normal state, trend warning state, restricted operation state, and emergency shutdown state. Corresponding control instructions are executed according to the operating status level. Continuous monitoring is maintained in normal status. Parameter records are reported and remote prompts are issued in trend warning status. In restricted operation status, the output power is reduced to the set threshold. In emergency shutdown status, the output circuit is immediately disconnected to ensure electrical insulation safety.

2. A photovoltaic inverter insulation resistance intelligent detection method according to claim 1, characterized in that: The bus voltage change data in the state transition stage obtains a continuous voltage value sequence through a preset sampling frequency. The voltage dynamic convergence modeling method includes calculating the voltage difference between adjacent sampling points and constructing a voltage change rate curve. The peak point and the starting point of the convergence stable interval are located on the curve, and the convergence time node is identified based on the condition that the voltage change rate is lower than the stability threshold in multiple consecutive sampling cycles; the start and end time intervals of the freezing window are dynamically calculated based on the time distance between the peak point and the convergence time node to automatically set the invalid interval of insulation resistance detection during shielding interference.

3. The photovoltaic inverter insulation resistance intelligent detection method according to claim 2, characterized in that: The resistance change trend result is obtained by performing a derivative calculation on the insulation resistance data within continuous sampling cycles. The calculation method is to divide the difference between the resistance value of the current cycle and the resistance value of the previous cycle by the length of the sampling cycle. The result is the resistance change rate. When the resistance change rate exceeds the set rate threshold for three or more consecutive sampling cycles, it is determined to be an abnormal trend state. The rate threshold is determined by the voltage resistance characteristics of the insulation material, the ambient temperature and humidity conditions and the historical data modeling results. It is set by any of the following methods: fixed empirical value, segmented dynamic value or logical interval based on fuzzy boundary setting.

4. A photovoltaic inverter insulation resistance intelligent detection method according to claim 3, characterized in that: The resistance change trend calibration behavior has a delayed confirmation mechanism. After identifying the initial abnormal trend, the mechanism starts two additional sampling cycles to continue observing the resistance change. If the resistance change rate in the new sampling cycles remains in the same direction and continues to exceed the aforementioned rate threshold, the trend is confirmed. If the rate in any new cycle is lower than the threshold or the change direction is opposite, the calibration behavior is canceled and the system returns to normal.

5. The photovoltaic inverter insulation resistance intelligent detection method according to claim 4, characterized in that: When the safety state decision process executes the weighted fusion path, a multivariate weighted scoring function is used to output the state level. This scoring function uses the resistance change trend results, ambient temperature, ambient humidity, load power, and grid voltage stability parameters as input variables. Each variable is uniformly normalized and converted into a score before input. The scoring function multiplies each score by a predefined weight factor and then sums them to form a state level score. This score is distributed between zero and one and is mapped to normal state, trend warning state, restricted operation state, or emergency shutdown state. The weight factors are configured through initialization parameters or generated through training of historical operation data.

6. A photovoltaic inverter insulation resistance intelligent detection method according to claim 5, characterized in that: Remote prompting in the trend warning state includes generating a status information package containing five types of data: the current insulation resistance change trend, the average ambient temperature and humidity over the past 30 seconds, the current load power, the grid voltage fluctuation amplitude, and the start and end time of the freeze window, and sending it to the remote monitoring platform through the data upload channel; The power reduction behavior under the restricted operating state includes calculating the power reduction target value. The power reduction target value is the product of the current output power and the power reduction factor. The power reduction factor is determined by the maximum change rate in the resistance change trend result and the current ambient temperature and humidity parameters. Its value is less than 1 and has multiple levels. If the grid voltage stability parameter is also lower than the stability threshold, the target power value will introduce a preset voltage attenuation compensation factor to perform an additional reduction.

7. The photovoltaic inverter insulation resistance intelligent detection method according to claim 6, characterized in that: Output circuit interruption control in the emergency stop state is achieved through parallel dual-path logic, where one path receives the emergency state signal output by the safety state decision process, and the other path receives auxiliary signal confirmation, including verification events where the resistance value is below the insulation safety lower limit for three consecutive cycles and the change rate direction is consistent. The disconnection operation is only performed when both paths are activated.

8. The photovoltaic inverter insulation resistance intelligent detection method according to claim 7, characterized in that: The start and end intervals of the freezing time window can be generated using different calculation methods based on preset strategies, including one of the following three methods: the first method is to collect the voltage change rate curve. When it is detected that the change rate is lower than the voltage stability threshold value within five consecutive sampling cycles, the interval is judged to be a voltage convergence segment. The starting point of the freezing window is the peak point of the curve, and the end point is the starting point of the convergence segment, forming a freezing interval between the peak point and the convergence point; the second method is to collect the maximum fluctuation value of the bus voltage within the voltage stability segment. When the fluctuation amplitude does not exceed the set offset threshold, the time interval corresponding to the continuous sampling cycles within the voltage stability segment is collected as the freezing window; the third method is to calculate the average convergence time based on historical operating data, and add a fixed safety margin as the window length, forming a freezing interval starting from the peak point.

9. The photovoltaic inverter insulation resistance intelligent detection method according to claim 8, characterized in that: The security status decision process consists of two structures: a rule-based decision path and a weighted fusion path. The rule-based decision path outputs the status level based on the conditional combination logic between the input variables. The weighted fusion path outputs the status level by weighted calculation of normalized score and weight factor; the two paths have a dynamic switching mechanism. The switching condition is that when the change amplitude of at least two types of input variables exceeds their respective stability thresholds at the same time during the continuous observation period, or when the variance of each variable exceeds the set composite standard, it will automatically switch from the rule judgment path to the weighted fusion path.

10. A photovoltaic inverter insulation resistance intelligent detection system, based on a photovoltaic inverter insulation resistance intelligent detection method according to any one of claims 1 to 9, characterized in that: It includes a state recognition module, an electrical modeling module, a trend assessment module, a safety decision module, and a remote interaction module. These modules work together to build an intelligent identification and protection response process for insulation performance, specifically including: The state recognition module is used to obtain the inverter operating status and bus voltage data in real time, and identify the state transition period corresponding to the startup and shutdown events; The electrical modeling module is used to perform convergence modeling on the bus voltage variation curve, extract the peak point and the starting point of the stable section, generate a frozen time window to shield the insulation resistance data within the period, and calculate the resistance change rate to construct a change trend curve; The trend assessment module is used to determine the resistance change rate over multiple consecutive sampling periods. If the threshold conditions are continuously met, it is marked as an abnormal trend and input into the subsequent safety determination process together with the ambient temperature and humidity, load power, and grid stability parameters. The safety decision module includes two decision structures: rule-based decision paths and weighted fusion paths. It also has a path switching mechanism that selects a decision path based on the fluctuation level of input variables and outputs four operating status levels: normal, trend warning, restricted operation, and emergency shutdown. The remote interaction module is used to generate a status data packet containing trend results, environmental information, electrical parameters and frozen window time period in the trend warning state or a higher level state, and send it to the remote monitoring platform to realize remote abnormality perception and system operation and maintenance linkage.

Images