Environment-friendly gas ring main unit intelligent control method for power supply stability regulation

Abstract

This invention discloses an intelligent control method for environmentally friendly gas ring main units (RNBs) used for power supply stability regulation, belonging to the field of power equipment control technology. It includes historical information acquisition: obtaining historical information such as electrical operating parameters, gas insulation status parameters, and insulation margin from the past operation of the environmentally friendly gas RNB. This invention acquires historical information from the past operation of the environmentally friendly gas RNB, extracts historical degradation fragments, establishes a historical insulation degradation sample library, and then converts the real-time state of the target environmentally friendly gas RNB into a state sequence that can characterize the current insulation change trend. This sequence is then matched with the historical degradation fragments to predict the insulation margin in future periods. Compared to relying solely on real-time monitoring values ​​for threshold alarms, this invention can identify the current stage of insulation evolution from historical degradation patterns, obtain the development trend of insulation risk in advance, and improve the foresight and accuracy of insulation status assessment.

Description

Technical Field

[0001] This invention relates to the field of power equipment control technology, specifically to an intelligent control method for an environmentally friendly gas ring main unit used for power supply stability regulation. Background Technology

[0002] Environmentally friendly gas-insulated ring main units are a new type of medium-voltage switchgear that uses environmentally friendly gases such as dry air, nitrogen, or perfluoroisobutyronitrile (PFOS) as insulation and arc-quenching media. The use of gases effectively reduces greenhouse gas emissions. With its compact internal structure, it achieves miniaturization, maintenance-free operation, and high reliability through a fully sealed design and gas insulation technology. It is particularly suitable for scenarios with high environmental and safety requirements, such as urban power distribution networks, new energy grid connection, and underground integrated pipe corridors.

[0003] A remote control method for an environmentally friendly gas ring main unit, disclosed in patent publication number CN117914008A, includes: Step S1, wirelessly transmitting monitored temperature, humidity, and air pressure parameters to an external terminal device via a central control module; Step S2, observing and inspecting various components within the ring main unit using a monitoring device; Step S3, when the central control module determines that the air pressure stability within the environmentally friendly gas ring main unit is below the allowable range, the central control module drives the air pump of the environmentally friendly gas ring main unit to operate. After the air pump operates, the air pressure is continuously monitored. If it is determined that the stability of the gas ring main unit falls below the allowable range again for more than 1 minute, the central control module sends a signal to the external terminal device to issue a warning and remind maintenance; Step S4, when it is determined that the temperature and humidity within the environmentally friendly gas ring main unit are abnormal, the central control module drives the heat dissipation device and dehumidification device within the environmentally friendly gas ring main unit to handle the situation.

[0004] In existing technologies, the operation control of environmental gas ring main units typically relies on real-time electrical quantities such as current, voltage, and switch status. It depends heavily on static threshold alarms or passive handling after a fault occurs, lacking mechanisms for constructing insulation samples based on historical degradation processes, matching similar degradation patterns, and predicting future insulation margins. Consequently, it is difficult to identify the evolution trend of insulation risks in a timely manner. Simultaneously, existing power supply regulation strategies often emphasize voltage qualification rate, power restoration, and network loss control, failing to incorporate insulation safety as a pre-emptive constraint into control decisions. This can easily lead to situations where electrical operating indicators improve but insulation risks intensify. Furthermore, for risks caused by changes in gas state, operating load, or switch actions, existing technologies generally lack risk source identification and graded proactive defense measures, resulting in insufficient targeting of regulation measures and difficulty in achieving early suppression of insulation risks and controlled operation. Therefore, this invention is proposed. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent control method for environmentally friendly gas ring main units for power supply stability regulation, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an intelligent control method for an environmentally friendly gas ring main unit for power supply stability regulation, the method comprising: Historical information acquisition: Obtain historical information by acquiring electrical operating parameters, gas insulation status parameters, and insulation margins from the past operation of the environmental protection gas ring main unit; Sample library establishment: Based on historical information, historical degradation fragments in the insulation degradation process are extracted through sample construction methods, and an insulation degradation historical sample library is established to store historical degradation fragments and subsequent insulation evolution data, and to establish the correspondence between historical degradation fragments and subsequent insulation evolution data; Real-time status acquisition: Determine the target environmentally friendly gas ring main unit, and acquire the real-time gas insulation status parameters and real-time electrical operation parameters of the target environmentally friendly gas ring main unit. The target environmentally friendly gas ring main unit corresponds to the environmentally friendly gas ring main unit that generated historical information. The real-time gas insulation status parameters are converted into a state sequence that can characterize the current insulation change trend through the state characterization method to obtain the current state sequence. Insulation trend prediction: Preset future time period, based on the current state sequence and the insulation degradation history sample library, the trend prediction method is used to perform similar degradation matching on the current state sequence and determine the historical degradation segment to obtain the target degradation segment. Based on the target degradation segment and the insulation degradation history sample library and corresponding relationship, the margin extrapolation method is used to calculate the insulation margin prediction value in the future time period to obtain the prediction result. Adjustment command generation: Based on real-time electrical operating parameters and prediction results, a power supply stability adjustment command is generated through a collaborative control method. The collaborative control method takes voltage deviation and network loss as optimization targets and insulation margin prediction value as a safety constraint. Active defense execution: Real-time monitoring of prediction results. When the predicted insulation margin value is detected to be lower than the preset insulation margin threshold, active defense instructions are generated and executed through active defense methods to bring the target environmental gas ring main unit into a state of controlled insulation risk operation.

[0007] Furthermore, the sample construction method includes: extracting a sequence of insulation margin changes over time based on historical information to obtain a historical insulation margin sequence; performing trend analysis based on the historical insulation margin sequence to identify various time segments in the historical insulation margin sequence where the insulation margin continuously decreases and the decrease exceeds the fluctuation range of normal equipment operation to obtain historical degradation process segments; slicing the historical degradation process segments using an adaptive time window to obtain several windows and the trace moisture content sequence, air pressure sequence, and temperature sequence within the windows; each window and the trace moisture content sequence, air pressure sequence, and temperature sequence within the window constitute a single historical degradation segment; extracting a historical insulation margin sequence of a fixed duration after the end time of the historical degradation segment to obtain subsequent insulation evolution data corresponding to the historical degradation segment; and establishing a historical insulation degradation sample to store all historical degradation segments and their corresponding subsequent insulation evolution data.

[0008] Furthermore, the length of the adaptive time window is determined based on the rate of decrease in insulation margin and the fluctuation inertia time of gas state parameters within the historical degradation process segment, with the fixed duration corresponding to the length of the adaptive time window.

[0009] Furthermore, the state characterization method includes: preprocessing the real-time gas insulation state parameters and real-time electrical operation parameters to obtain real-time state basic data for time alignment, missing value repair, and noise reduction; extracting the rate of change of micro-moisture content, gas pressure, temperature, and insulation margin based on the real-time state basic data; eliminating the influence of operating conditions on the rate of change of insulation margin by combining the load current, voltage fluctuation, and switching action information in the real-time electrical operation parameters to obtain the de-operating condition disturbance state component; arranging the de-operating condition disturbance state components at multiple consecutive moments in time sequence based on the sampling step size; constructing a multi-dimensional state vector that can characterize the insulation change trend based on the parameter change direction, change amplitude, and duration at each moment; and combining the multi-dimensional state vectors at multiple consecutive moments in time sequence to obtain the current state sequence for subsequent similarity degradation matching.

[0010] Furthermore, the trend prediction method includes: aligning the current state sequence with each historical degradation segment in the insulation degradation historical sample library according to the same parameter order and the same time order; comparing the direction of change, magnitude of change, order of occurrence of continuous change time periods, and primacy relationship of corresponding parameters in the current state sequence and each historical degradation segment at corresponding times; eliminating historical degradation segments where corresponding parameters have opposite directions of change at the same corresponding time, different order of occurrence of continuous change time periods, or different primacy relationships; based on the historical degradation segments retained after elimination, accumulating the difference in magnitude of change of each corresponding parameter at each corresponding time in chronological order, and retaining the historical degradation segment with the smallest accumulation result to obtain one or more target degradation segments; when multiple target degradation segments are obtained, using all of the multiple target degradation segments as input for subsequent margin extrapolation; the parameter order includes the rate of change of trace moisture content, the rate of change of air pressure, the rate of change of temperature, and the rate of change of insulation margin.

[0011] Furthermore, the margin estimation method includes: retrieving subsequent insulation evolution data corresponding to the target degradation segment based on the historical insulation degradation sample library, extracting the subsequent insulation margin sequence arranged in chronological order, determining the actual insulation margin of the target environmental gas ring main unit corresponding to the end time of the current state sequence as the predicted starting insulation margin, calculating the insulation margin change of each subsequent time in the corresponding subsequent insulation margin sequence relative to the starting time when there is only one target degradation segment, and applying it sequentially to the predicted starting insulation margin to obtain the initial insulation margin prediction value, and calculating the insulation margin change of each subsequent time in the subsequent insulation margin sequence corresponding to each target degradation segment relative to its respective starting time when there are multiple target degradation segments, selecting the change in the middle position of multiple changes at the same time position as the representative change. The initial insulation margin prediction value is obtained by applying the quantity to the insulation margin prediction starting point in chronological order. The load current change direction sequence, voltage fluctuation direction sequence, and switch action sequence within the time period corresponding to the current state sequence of the target environmental gas ring main unit are extracted and compared with the corresponding sequences within the historical time period corresponding to each target degradation segment. When the comparison results of each sequence are the same, the initial insulation margin prediction value is used as the insulation margin prediction value. When the comparison results of each sequence are different, the target degradation segment that is inconsistent with the current state sequence is removed, and the insulation margin prediction value is recalculated based on the retained target degradation segment. If no target degradation segment is retained after removal, the initial insulation margin prediction value obtained before removal is retained, and the insulation margin prediction value is obtained by extracting the prediction values ​​of each moment corresponding to the future time period based on the time series of the initial insulation margin prediction value.

[0012] Furthermore, the collaborative control method includes: acquiring the open / closed status of each switch, the on / off relationship of each line, the node voltage of each node, the line current of each line, the load access relationship of each load access point, and the activation status of each reactive power compensation device within the power supply area where the target environmental gas ring network cabinet is located; determining the current power supply path based on the open / closed status of each switch and the on / off relationship of each line; determining the load range supplied by the target environmental gas ring network cabinet based on the load access relationship of each load access point; acquiring candidate control actions that can change the current power supply path or load range, the candidate control actions including sectional switch opening, sectional switch closing, tie switch opening, tie switch closing, load switching, and reactive power compensation device activation / deactivation; performing deduction on each candidate control action to obtain the power supply path, load range, node voltage, line current, and line loss after execution; and eliminating actions that fail to execute the desired control actions. If a candidate control action results in the loss of power to the previously supplied load, an expansion of the load range borne by the target environmental gas ring main unit, or an insulation margin prediction value lower than the insulation margin threshold, then the candidate control action that reduces the load range borne by the target environmental gas ring main unit is prioritized. If multiple candidate control actions exist, the candidate control action that reduces the node voltage deviation after execution is retained. If multiple candidate control actions still exist after retention, the candidate control action that reduces line loss after execution is retained. The retained candidate control actions are then combined and combined in sequence. The combination of control actions that simultaneously satisfies the following conditions is selected: the previously supplied load is kept powered; the load range borne by the target environmental gas ring main unit is not expanded; the insulation margin prediction value is not lower than the insulation margin threshold; the node voltage deviation is reduced; and the line loss is reduced. This combination is then converted into a power supply stabilization adjustment command according to the action execution order.

[0013] Furthermore, the active defense method includes: checking the predicted insulation margin values ​​at each moment in chronological order; determining the moment when the predicted insulation margin value first falls below the insulation margin threshold to obtain the target moment; defining the target moment and subsequent periods below the insulation margin threshold as risk periods; extracting the current state sequence before the start of the risk period, as well as the real-time gas insulation state parameters and real-time electrical operation parameters within the risk period to form pre-risk gas state records, pre-risk operation state records, gas state records during the risk period, and operation state records during the risk period, respectively; and comparing the pre-risk gas state records with the gas state records during the risk period to determine the direction of change in moisture content, gas pressure, and temperature. The risk source is determined by comparing the pre-risk and in-risk operating status records to assess whether the direction of load current change, voltage fluctuation direction, and switch operation sequence have changed. If at least one of the following changes—moisture content change direction, gas pressure change direction, or temperature change direction—has changed, and none of the load current change direction, voltage fluctuation direction, or switch operation sequence has changed, the risk source is determined to be a change in gas insulation condition. If either the load current change direction or voltage fluctuation direction has changed, and none of the moisture content change direction, gas pressure change direction, or temperature change direction has changed, the risk source is determined to be a change in operating load. If the switch operation sequence has changed... If the change occurs before the start of the risk period or during the duration of the risk period, the risk source is determined to be a change in switch operation. When two or three of the above risk sources are simultaneously present, an active defense command is generated in the following order: first, reduce the load on the target environmental gas ring main unit; then, stop all switch operations except those necessary to maintain the current power supply path connection; and finally, implement load switching. When the risk source is a change in gas insulation status, an active defense command is generated to reduce the load on the target environmental gas ring main unit and prohibit the transfer of new loads. When the risk source is a change in operating load, a command is generated to switch the load whose load current change direction has changed from the target environmental gas ring main unit to an adjacent power supply path. The active defense command is generated when the risk source is a change in switch operation. It stops all switch operations except those necessary to maintain the current power supply path connection and keeps the current power supply path unchanged. After executing the active defense command, the insulation margin prediction value for the subsequent period is re-acquired. When the re-acquired insulation margin prediction value is no longer lower than the insulation margin threshold, the current operating state is maintained. When the re-acquired insulation margin prediction value is still lower than the insulation margin threshold, the active defense command to reduce the load on the target environmental gas ring main unit is continued to be executed. When the re-acquired insulation margin prediction value continues to decrease, the active defense command to make the target environmental gas ring main unit withdraw from load transfer or withdraw from operation is executed.

[0014] Compared with the prior art, the beneficial effects of the present invention are: This intelligent control method for environmentally friendly gas ring main units (RMNs) used for power supply stability regulation first acquires electrical operating parameters, gas insulation status parameters, and insulation margin from the RMN's past operation. It then extracts historical degradation segments during the insulation degradation process and establishes a historical insulation degradation sample library. Next, it converts the real-time state of the target RMN into a state sequence characterizing the current insulation change trend and performs similar degradation matching with the historical degradation segments, thereby predicting the insulation margin for future periods. Compared to threshold alarms based solely on real-time monitoring values, this invention can identify the current stage of insulation evolution from historical degradation patterns, obtain the development trend of insulation risks in advance, and improve the foresight and accuracy of insulation status assessment.

[0015] Simultaneously, the predicted insulation margin is further incorporated as a safety constraint into the power supply stability regulation process. When generating regulation commands, not only are power supply operation indicators such as voltage deviation and network loss considered, but the future insulation safety level of the target environmentally friendly gas ring main unit is also simultaneously constrained. By deduce, screen, and optimize candidate control actions and their combinations, load transfer, power supply path adjustment, and reactive power compensation coordination control can be prioritized without expanding the load range of risky equipment. This balances power supply continuity, node voltage quality, line loss, and insulation safety, thereby improving the comprehensive optimization capability of ring main unit operation control.

[0016] Simultaneously, an active defense mechanism based on prediction results is established. When the predicted insulation margin is detected to be lower than a preset threshold, the mechanism can differentiate and judge the source of risk by combining changes in gas state and operating state before and after the risk, and generate targeted active defense instructions according to different risk sources. This not only enables timely reduction of the load on the target ring main unit, restriction of unnecessary switching actions, or implementation of load switching before the insulation risk expands, but also allows for re-predicting the insulation margin and continuously revising the control strategy after the defensive measures are implemented, so that the target environmental gas ring main unit enters a controlled operating state of insulation risk, enhancing the system's proactive handling capability and operational reliability in the face of insulation degradation risks. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall process of the present invention; Figure 2 This is a schematic diagram of the sample library establishment method of the present invention; Figure 3 This is a schematic diagram of the insulation trend prediction process of the present invention; Figure 4 This is a schematic diagram of the collaborative control generation and adjustment instruction flow of the present invention; Figure 5 This is a schematic diagram of the active defense execution process of the present invention. Detailed Implementation

[0018] 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.

[0019] The environmentally friendly gas-insulated ring main unit (RMU) for power supply stability regulation deeply integrates intelligent control technologies such as feeder automation, fault self-healing, and load transfer into the traditional RMU power distribution function. When a short circuit or ground fault occurs in the distribution network, it can quickly isolate the faulty section using its built-in intelligent terminal and communication module, and automatically switch the tie switch to transfer power to the backup power source for the non-faulty area. The entire process of self-healing and reconstruction is completed in milliseconds, significantly reducing power outage time for users. Simultaneously, thanks to the excellent insulation performance of the environmentally friendly gas, the device can maintain long-term stable operation in various harsh environments. Combined with distributed energy storage interfaces and dynamic reactive power compensation functions, it effectively smooths voltage fluctuations and load impacts, thereby comprehensively improving the power grid's power supply reliability, power quality, and adaptive regulation capabilities. Example 1:

[0020] like Figures 1-5 As shown, the present invention provides a technical solution: an intelligent control method for an environmentally friendly gas ring main unit for power supply stability regulation, the method comprising: Historical information acquisition: Obtain historical information by acquiring electrical operating parameters, gas insulation status parameters, and insulation margins from the past operation of the environmental protection gas ring main unit; It is important to note that historical information should include at least electrical operating parameters, gas insulation status parameters, and insulation margin. Historical information can be sourced from dispatch automation systems, distribution terminals, ring main unit online monitoring devices, partial discharge monitoring modules, micro-moisture monitoring modules, temperature and pressure acquisition modules, or station-level data platforms. To ensure comparability between different data sources, historical information can be uniformly converted into time-series data with the same sampling period. Electrical operating parameters include load current, voltage fluctuation information, switch operation information, the open / closed status of each switch, voltage at each node, current in each line, the on / off relationship of each line, the load connection relationship at each load connection point, and the operational status of each reactive power compensation device. Gas insulation status parameters include micro-moisture content, gas pressure, and temperature, and may further include one or more of the following: gas density, dew point, leakage status, partial discharge characteristics, and gas component ratio.

[0021] Sample library establishment: Based on historical information, historical degradation fragments in the insulation degradation process are extracted through sample construction methods, and an insulation degradation historical sample library is established to store historical degradation fragments and subsequent insulation evolution data, and to establish the correspondence between historical degradation fragments and subsequent insulation evolution data; It is important to note that the historical insulation degradation sample library stores multiple historical degradation segments and corresponding subsequent insulation evolution data for each segment. The historical degradation segments characterize a trend in the insulation state of the environmentally friendly gas ring main unit before insulation degradation occurs or accelerates, while the subsequent insulation evolution data characterizes the trajectory of insulation margin changes after the end of the historical degradation segment. By establishing these correspondences, the current equipment state, after matching its similarity to a historical segment, can directly access the corresponding subsequent historical trajectory to extrapolate future insulation margins.

[0022] Real-time status acquisition: Determine the target environmentally friendly gas ring main unit, and acquire the real-time gas insulation status parameters and real-time electrical operation parameters of the target environmentally friendly gas ring main unit. The target environmentally friendly gas ring main unit corresponds to the environmentally friendly gas ring main unit that generated historical information. The real-time gas insulation status parameters are converted into a state sequence that can characterize the current insulation change trend through the state characterization method to obtain the current state sequence. It is important to note that the current stage is the actual operation stage. During the real-time operation stage, the target environmentally friendly gas ring main unit is identified, and its real-time gas insulation status parameters and real-time electrical operation parameters are continuously acquired. These real-time parameters are then converted into a current state sequence using a state characterization method. This current state sequence is not a simple concatenation of the original sampled values, but rather a state expression representing the direction, rate, and duration of parameter changes, as well as the trend information after the elimination of the influence of the operating conditions. This avoids misjudgments due to random fluctuations caused by relying solely on the absolute parameter values ​​at a single moment. By correlating the target environmentally friendly gas ring main unit with those that generated historical information, it ensures that the information in the insulation degradation historical sample database can be applied to the target environmentally friendly gas ring main unit.

[0023] Insulation trend prediction: Preset future time period, based on the current state sequence and the insulation degradation history sample library, the trend prediction method is used to perform similar degradation matching on the current state sequence and determine the historical degradation segment to obtain the target degradation segment. Based on the target degradation segment and the insulation degradation history sample library and corresponding relationship, the margin extrapolation method is used to calculate the insulation margin prediction value in the future time period to obtain the prediction result. It is important to note that after constructing the current state sequence, it is compared with each historical degradation segment in the insulation degradation history sample library. The target degradation segment that is closest to the current state is selected. Then, based on the target degradation segment and its corresponding subsequent insulation evolution data, the insulation margin of the target environmental gas ring main unit in the future period is extrapolated to obtain the insulation margin prediction value. The future period can be set to several minutes, tens of minutes, several hours or longer in the future according to the actual power distribution operation needs.

[0024] Regulation command generation: Based on real-time electrical operating parameters and prediction results, power supply stability regulation commands are generated through a collaborative control method. The collaborative control method takes voltage deviation and network loss as optimization targets and insulation margin prediction value as safety constraints. It is important to note that after obtaining the insulation margin prediction value, the power supply stability adjustment command is generated by further considering the current power supply task undertaken by the target environmental gas ring main unit, the regional power supply topology, and the execution effect of candidate adjustment actions. In this process, voltage deviation and network loss are used as optimization targets, while ensuring that the predicted insulation margin value is not lower than the insulation margin threshold is used as a safety constraint. That is, if an adjustment action can improve voltage deviation or reduce network loss, but would cause the future predicted insulation margin value of the target environmental gas ring main unit to fall below the safety threshold, then that adjustment action will not be adopted.

[0025] Active defense execution: Real-time monitoring of prediction results. When the predicted insulation margin value is detected to be lower than the preset insulation margin threshold, active defense instructions are generated and executed through active defense methods to bring the target environmental gas ring main unit into a state of controlled insulation risk operation.

[0026] It should be noted that, in addition, when the predicted insulation margin value is detected to be lower than the insulation margin threshold, the system will no longer only perform routine power supply adjustments, but will further enter the active defense mode. The active defense mode is used to analyze the source of risk and generate defense commands including load reduction, stopping unnecessary switching actions, load switching or shutdown, so as to bring the target environmental gas ring main unit into an insulation risk controlled operation state. The insulation margin threshold is determined according to the actual situation.

[0027] For example, a ring main unit using environmentally friendly gas insulation in a certain power supply area has accumulated historical data over the past two years, including moisture content, air pressure, temperature, load current, voltage fluctuations, and insulation margin. The system uses this historical data to establish a historical sample library of insulation degradation. During operation on a certain day, the target ring main unit experiences abnormal fluctuations in insulation parameters due to rising temperature and load fluctuations. The system collects real-time data on the rising trend of moisture content, the slow decreasing trend of air pressure, the rising trend of temperature, and current fluctuations, converting these into a current state sequence. Subsequently, the system matches several historical degradation segments from the historical sample library that are similar to the current state trend, further predicting that the insulation margin will continue to decrease within the next 30 minutes and fall below the insulation margin threshold at the 22nd minute. Based on this prediction, the system prioritizes switching some loads to adjacent power supply paths among the candidate control actions to reduce the burden on the ring main unit. If, after execution, the prediction still indicates that the insulation margin will continue to decrease, active defense is further triggered, restricting new load access and stopping unnecessary switching operations, thereby preventing the expansion of insulation risks.

[0028] like Figure 1 and Figure 2 As shown, the sample construction method includes: extracting the insulation margin variation sequence over time based on historical information to obtain a historical insulation margin sequence; performing trend analysis based on the historical insulation margin sequence to identify each time segment in the historical insulation margin sequence where the insulation margin continuously decreases and the decrease exceeds the fluctuation range of normal equipment operation to obtain a historical degradation process segment; slicing the historical degradation process segment with an adaptive time window to obtain several windows and the micro-moisture content sequence, air pressure sequence, and temperature sequence within the window; a single window and the micro-moisture content sequence, air pressure sequence, and temperature sequence within the window constitute a single historical degradation segment; extracting the historical insulation margin sequence of a fixed duration after the end time of the historical degradation segment to obtain the subsequent insulation evolution data corresponding to the historical degradation segment; and establishing an insulation degradation historical sample to store all historical degradation segments and their corresponding subsequent insulation evolution data.

[0029] It is important to note that the sample construction method is used to extract historical segments with degradation indication significance from historical information, rather than treating all historical data equally. Therefore, firstly, a historical insulation margin sequence is extracted based on historical information to show how insulation margin changes over time. This historical insulation margin sequence can be arranged according to a uniform sampling period, such as sampling once every 1 minute, 5 minutes, or 15 minutes. Then, trend analysis is performed on the historical insulation margin sequence to identify time periods where the insulation margin continuously decreases and the decrease exceeds the normal operating fluctuation range of the equipment. The normal operating fluctuation range of the equipment can be obtained based on the typical operating condition fluctuation range given by the equipment manufacturer, the statistical fluctuation range of historical healthy operation phases, or a safety fluctuation range set based on experience. When the insulation margin continuously decreases within a certain time period, and the cumulative decrease exceeds the above-mentioned normal fluctuation range, this time period can be identified as having degradation process characteristics and marked as a historical degradation process segment. After identifying the historical degradation process segment, each historical... The historical degradation process segment is sliced ​​using an adaptive time window. Instead of mechanically truncating with a fixed-length window, multiple windows are extracted based on the continuous characteristics of the insulation degradation process. For each window, the corresponding moisture content sequence, pressure sequence, and temperature sequence are extracted. A single window and its corresponding moisture content sequence, pressure sequence, and temperature sequence together constitute a historical degradation segment. This method preserves the local features of parameter changes before and after degradation. After obtaining the historical degradation segments, the historical insulation margin sequence within a fixed time after the end of each historical degradation segment is further extracted as the subsequent insulation evolution data corresponding to that historical degradation segment. The subsequent insulation evolution data reflects whether the insulation margin continues to decline, tends to stabilize, or partially recovers after the historical degradation segment. Finally, all historical degradation segments and their corresponding subsequent insulation evolution data are stored in the historical insulation degradation sample library for subsequent similarity matching and trend prediction.

[0030] For example, analyzing one year's historical data of a target type of environmentally friendly gas ring main unit yields its insulation margin time series. The system discovers that the insulation margin in several time segments continuously decreases within 2 hours, and the total decrease significantly exceeds the daily fluctuation range of the healthy operation phase. These time segments are then identified as historical degradation process segments. Subsequently, the system uses adaptive window slicing for each historical degradation process segment. For example, in a certain degradation process, multiple windows with lengths of 20 minutes, 25 minutes, and 30 minutes are segmented, and the corresponding trace moisture content sequence, gas pressure sequence, and temperature sequence are extracted within each window. Then, the insulation margin sequence for the next 30 minutes after the end of each window is extracted as its subsequent insulation evolution data. In the sample library established in this way, each sample contains "degradation precursor segments + subsequent insulation change trajectory", which facilitates subsequent analogy and deduction.

[0031] like Figure 1As shown, the length of the adaptive time window is determined based on the rate of decrease of the insulation margin and the fluctuation inertia time of the gas state parameters during the historical degradation process. The fixed duration corresponds to the length of the adaptive time window.

[0032] It is important to note that the length of the adaptive time window is not fixed in advance, but is determined jointly based on the rate of decrease of insulation margin within the historical degradation process and the fluctuation inertia time of the gas insulation state parameters. The rate of decrease of insulation margin refers to the magnitude of change in insulation margin per unit time; the fluctuation inertia time refers to the typical time length during which parameters such as moisture content, gas pressure, and temperature, under certain operational disturbances or environmental changes, deviate from their initial deviation to the formation of a stable trend. When the insulation margin decreases rapidly within the historical degradation process, it indicates a more sensitive degradation evolution, and the window length can be correspondingly reduced to highlight the short-term rapid change characteristics. When the insulation margin decreases slowly but the gas state parameter changes exhibit strong inertia, the window length can be appropriately increased to preserve the complete trend formation process. The fixed duration corresponds to the adaptive time window length and can be set to be the same as the window length, or set to a duration corresponding to a predetermined proportion, thereby maintaining consistency between the historical degradation segment and subsequent insulation evolution data on the time scale.

[0033] For example, in one type of historical degradation process, the insulation margin decreases rapidly within 10 minutes, and the moisture content and air pressure show significant coordinated changes within the first 8 minutes. In this case, the adaptive window length can be set to 10 minutes or 12 minutes to preserve key precursors to degradation. In another type of historical degradation process, the insulation margin decreases more slowly, and the changes in moisture content and temperature take about 25 minutes to form a clear trend. In this case, the adaptive window length can be set to 25 minutes or 30 minutes. Correspondingly, the extraction time of subsequent insulation evolution data can be consistent with the window length. For example, if the window is 25 minutes, the insulation margin sequence 25 minutes after the end of the extraction segment can be used as the subsequent insulation evolution data.

[0034] like Figure 1 and Figure 3As shown, the state characterization method includes: preprocessing the real-time gas insulation state parameters and real-time electrical operation parameters to obtain real-time state basic data for time alignment, missing value repair, and noise reduction; extracting the rate of change of micro-moisture content, gas pressure, temperature, and insulation margin based on the real-time state basic data; and eliminating the influence of operating conditions on the rate of change of insulation margin by combining the load current, voltage fluctuation, and switching action information in the real-time electrical operation parameters to obtain the de-operating condition disturbance state component; arranging the de-operating condition disturbance state components at multiple consecutive time points in a temporal sequence based on the sampling step size; and constructing a multi-dimensional state vector that can characterize the insulation change trend based on the parameter change direction, change amplitude, and duration at each time point; and combining the multi-dimensional state vectors at multiple consecutive time points in a temporal order to obtain the current state sequence for subsequent similarity degradation matching.

[0035] It is important to note that the state characterization method is used to convert the real-time raw data of the target environmental gas ring main unit into a structured state sequence that can be used for trend matching. First, the real-time gas insulation state parameters and real-time electrical operation parameters are preprocessed with time alignment, missing value repair, and denoising. Time alignment eliminates the impact of asynchronous sampling times from different sensors; missing value repair fills in data gaps caused by short-term communication interruptions or missing sampling; denoising removes random jitter, transient spikes, or inaccurate changes. After preprocessing, the rates of change in moisture content, gas pressure, temperature, and insulation margin are extracted based on the real-time state data. Using rates of change instead of raw values ​​helps highlight the trend of parameter changes and improves the consistency of comparisons across devices and time periods. Furthermore, by combining the load current, voltage fluctuations, and switching action information from the real-time electrical operation parameters, the operational condition influences on the rates of change in moisture content, gas pressure, temperature, and insulation margin are eliminated. The aim is to minimize non-degradable changes caused by short-term load fluctuations, transient voltage disturbances, or switching operations, making the remaining components more representative of the true trend of the equipment's insulation state. After eliminating the influence of operating conditions, the disturbance-free state components from multiple consecutive time points are sequentially arranged according to the sampling step size, and a multi-dimensional state vector is constructed based on the direction, amplitude, and duration of parameter changes at each time point. For example, the direction of change can be represented as increasing, decreasing, or remaining essentially unchanged; the amplitude of change can be represented as the increment between adjacent sampling points; and the duration can represent the number of sampling points maintained in a certain direction of change. Finally, the multi-dimensional state vectors from multiple consecutive time points are combined in chronological order to obtain the current state sequence, which is used for subsequent similarity matching with historical degradation segments.

[0036] For example, a target environmental protection gas ring main unit samples data every minute for the past 15 minutes, obtaining data on moisture content, gas pressure, temperature, load current, and insulation margin. The system first aligns the different sampling sources in time, interpolates missing sampling points, and then uses a sliding smoothing method to remove high-frequency noise. Afterward, it calculates the rates of change for moisture content, gas pressure, temperature, and insulation margin. Considering the short-term load surge and a switching operation at the 5th and 6th minutes, the system uses load current and switching action information to eliminate the impact of operating conditions on parameter changes at the corresponding time points. After processing, the system arranges the state components of the most recent 15 sampling moments in chronological order, constructing a current state sequence containing the direction, magnitude, and duration of change, which is then used for matching with a historical sample database.

[0037] like Figure 1 and Figure 3 As shown, the trend prediction method includes: aligning the current state sequence with each historical degradation segment in the insulation degradation historical sample library according to the same parameter order and the same time order; comparing the direction of change, magnitude of change, order of occurrence of continuous change time periods, and primacy relationship of the corresponding parameters in the current state sequence and each historical degradation segment at the corresponding time; eliminating historical degradation segments where the corresponding parameters change in opposite directions at the same corresponding time, have different order of occurrence of continuous change time periods, or have different primacy relationships; based on the historical degradation segments retained after elimination, accumulating the difference in magnitude of change of each corresponding parameter at each corresponding time in chronological order, and retaining the historical degradation segment with the smaller or smallest accumulated result to obtain one or more target degradation segments; when multiple target degradation segments are obtained, all of the multiple target degradation segments are used as inputs for subsequent margin extrapolation, and the parameter order includes the rate of change of micro-moisture content, the rate of change of air pressure, the rate of change of temperature, and the rate of change of insulation margin.

[0038] It is important to note that the trend prediction method compares the current state sequence with historical degradation segments in the insulation degradation historical sample library to select the target degradation segment that best matches the current degradation trend. To ensure consistency in the comparison, the current state sequence and each historical degradation segment are first aligned according to the same parameter order and the same time order. The parameter order includes the rate of change of trace moisture content, the rate of change of air pressure, the rate of change of temperature, and the rate of change of insulation margin. After alignment, the direction of change, the magnitude of change, the order of occurrence of continuous change time periods, and the primacy relationship relative to the rate of change of insulation margin are compared between the corresponding parameters in the current state sequence and each historical degradation segment at the corresponding time. The primacy relationship is used to characterize temporal features such as "the change of trace moisture content precedes the change of insulation margin" and "the change of air pressure lags behind the change of insulation margin," thereby improving the accuracy of the comparison. The ability to identify similarities in degradation mechanisms; the order of occurrence of the continuous change time periods is used to characterize the temporal structure of each parameter continuously rising, continuously falling, or continuously stabilizing over a period of time; if a historical degradation segment shows a change direction opposite to the current state at the same corresponding moment, or if the order of occurrence of its continuous change time periods is inconsistent with the current state, or if the precedence relationship of its parameters relative to the rate of change of insulation margin is inconsistent with the current state, it is considered to be inconsistent with the current degradation mechanism and is removed; for the historical degradation segments that are retained after removal, the difference in the change amplitude of each corresponding parameter at each corresponding moment is accumulated in chronological order, and the historical degradation segments with smaller or smaller accumulated results are retained to obtain one or more target degradation segments; this method of "first rule screening, then amplitude fine screening" can take into account both the consistency of degradation patterns and the closeness of numerical values.

[0039] For example, the current state sequence is characterized by: continuous increase in trace moisture content, slow decrease in air pressure, initial rise followed by stabilization in temperature, and continuous decrease in insulation margin. Furthermore, the change in trace moisture content precedes the change in insulation margin by two sampling steps, and the change in air pressure precedes the change in insulation margin by one sampling step. The system compares this current state sequence with multiple historical degradation segments in the sample library one by one. If the air pressure in a historical degradation segment shows an upward trend at the same time, it is removed because the direction is opposite. If another historical degradation segment shows the same direction, but its temperature change occurs after the insulation margin change, while the temperature change in the current state occurs before the insulation margin change, it is removed because the order of occurrence is different. For the remaining segments, the difference in the magnitude of change of each parameter at each time point is calculated and accumulated over time. One or more segments with the smallest accumulated result are selected as the target degradation segments.

[0040] like Figure 1 and Figure 3As shown, the margin extrapolation method includes: retrieving subsequent insulation evolution data corresponding to the target degradation segment from the historical insulation degradation sample library, extracting the subsequent insulation margin sequence arranged in chronological order, determining the actual insulation margin of the target environmental gas ring main unit corresponding to the end time of the current state sequence as the predicted starting insulation margin, calculating the insulation margin change of each subsequent time in the corresponding subsequent insulation margin sequence relative to the starting time when there is only one target degradation segment, and applying it sequentially to the predicted starting insulation margin to obtain the initial insulation margin prediction value, when there are multiple target degradation segments, calculating the insulation margin change of each subsequent time in the subsequent insulation margin sequence corresponding to each target degradation segment relative to its respective starting time, selecting the change in the middle position of multiple changes at the same time position as the representative change, and... The initial insulation margin prediction value is obtained by sequentially applying the predicted starting insulation margin in chronological order. The load current change direction sequence, voltage fluctuation direction sequence, and switch action sequence within the time period corresponding to the current state sequence of the target environmental gas ring main unit are extracted and compared with the corresponding sequences within the historical time period corresponding to each target degradation segment. When the comparison results of each sequence are the same, the initial insulation margin prediction value is used as the insulation margin prediction value. When the comparison results of each sequence are different, the target degradation segment that is inconsistent with the current state sequence is removed, and the insulation margin prediction value is recalculated based on the retained target degradation segment. If no target degradation segment is retained after removal, the initial insulation margin prediction value obtained before removal is retained, and the insulation margin prediction value is obtained by extracting the prediction values ​​of each moment corresponding to the future time period based on the time series of the initial insulation margin prediction value.

[0041] It is important to note that the margin extrapolation method is used to predict the insulation margin of a target environmentally friendly gas ring main unit in future periods based on the historical subsequent evolution data corresponding to the target degradation segment. First, it retrieves the subsequent insulation evolution data corresponding to the target degradation segment and extracts the subsequent insulation margin sequence. Then, it uses the actual insulation margin of the target environmentally friendly gas ring main unit at the end of the current state sequence as the predicted starting insulation margin. When there is only one target degradation segment, it calculates the change in insulation margin at each subsequent time point relative to the starting time in the subsequent insulation margin sequence corresponding to that segment, and applies these changes sequentially to the predicted starting insulation margin to obtain the initial insulation margin prediction value for each future time point. Essentially, it shifts the insulation change trend after similar historical degradation segments to the current time point. When there are multiple target degradation segments, it calculates the change in the subsequent insulation margin sequence corresponding to each target degradation segment separately, and selects the change at the middle position among multiple changes at the same time point as the representative change, i.e., using the middle position. The method of suppressing the influence of extreme samples is used; then, the representative changes at each time position are applied to the insulation margin at the prediction starting point in chronological order to obtain the initial insulation margin prediction value; considering that although the historical degradation segment and the current state are similar in terms of insulation state changes, their operating conditions at that time may not be completely consistent with the current operating conditions, the load current change direction sequence, voltage fluctuation direction sequence, and switch action sequence are further extracted within the time period corresponding to the current state sequence of the target environmental gas ring main unit, and compared with the corresponding sequences within the historical time period corresponding to each target degradation segment; when the comparison results of each sequence are the same, it indicates that the historical operating conditions and the current operating conditions are highly consistent, and the initial insulation margin prediction value can be directly used as the final insulation margin prediction value; when there is inconsistency, the target degradation segment that is inconsistent with the current operating conditions is removed, and the insulation margin prediction value is recalculated based on the retained segment; if no target degradation segment is retained after removal, in order to avoid prediction interruption, the initial insulation margin prediction value obtained before removal can be output as the insulation margin prediction value.

[0042] For example: After trend matching, the system obtains three target degradation segments; taking the current actual insulation margin of 0.32 as the prediction starting point, the insulation margin changes corresponding to the three target degradation segments in the following 10 minutes are as follows: Sample A is {-0.01, -0.02, -0.03, ...} from minute 1 to minute 10, Sample B is {-0.01, -0.01, -0.025, ...}, and Sample C is {-0.015, -0.02, -0.028, ...}; the system analyzes multiple changes at each minute position. The intermediate values ​​are taken for the quantity, for example, -0.01 for the first minute, -0.02 for the second minute, and -0.028 for the third minute, thus obtaining the initial insulation margin prediction value sequence. Subsequently, the system compares the load current change direction sequence, voltage fluctuation direction sequence, and switching action sequence of the current time period with the historical time periods corresponding to the above three target degradation segments. If it is found that the historical time period corresponding to sample C has a switching action sequence that is inconsistent with the current one, sample C is removed, and the final insulation margin prediction value is recalculated based on sample A and sample B.

[0043] like Figure 1 and Figure 4 As shown, the collaborative control method includes: acquiring the open / closed status of each switch, the on / off relationship of each line, the node voltage of each node, the line current of each line, the load access relationship of each load access point, and the activation status of each reactive power compensation device within the power supply area where the target environmental gas ring network cabinet is located; determining the current power supply path based on the open / closed status of each switch and the on / off relationship of each line; determining the load range to be powered by the target environmental gas ring network cabinet based on the load access relationship of each load access point; acquiring candidate control actions that can change the current power supply path or load range, including sectional switch opening, sectional switch closing, tie switch opening, tie switch closing, load switching, and reactive power compensation device switching; performing a simulation on each candidate control action to obtain the power supply path, load range, node voltage, line current, and line loss after execution; and eliminating actions that would cause changes to the original power supply. For candidate control actions such as power loss of electrical load, expansion of the load range of the target environmental gas ring main unit, or insulation margin prediction value being lower than the insulation margin threshold, priority is given to retaining candidate control actions that reduce the load range of the target environmental gas ring main unit. When multiple candidate control actions exist, further retain candidate control actions that reduce node voltage deviation after execution. If multiple candidate control actions still exist after further retention, then retain candidate control actions that reduce line loss after execution. The retained candidate control actions are then combined and combined in sequence to select the control action combination that simultaneously satisfies the following conditions: maintaining power supply to the already powered load, not expanding the load range of the target environmental gas ring main unit, insulation margin prediction value not lower than the insulation margin threshold, and reducing node voltage deviation and line loss. This combination is then converted into a power supply stabilization adjustment command according to the action execution order.

[0044] It is important to note that the collaborative control method is used to generate power supply stability adjustment commands under insulation safety constraints. First, real-time operational information of the power supply area where the target environmental gas ring main unit is located is acquired, including the open / closed status of each switch, the on / off relationship of each line, the node voltage of each node, the line current of each line, the load connection relationship of each load connection point, and the operational status of each reactive power compensation device. Based on this information, the current power supply path and the load range supplied by the target environmental gas ring main unit are determined. Then, candidate control actions that can change the current power supply path or load range are listed, such as opening or closing sectionalizing switches, opening or closing tie switches, load switching, and switching on / off of reactive power compensation devices. Each candidate control action is simulated to obtain the power supply path, load range, node voltage, line current, and line loss after execution. The deduction here can be implemented based on power distribution network flow calculation, topology reconfiguration calculation, constraint verification model, or simulation module. In the candidate action screening stage, actions that would cause the already powered load to lose power after execution are first eliminated; actions that would expand the load range of the target environmental gas ring main unit are also eliminated; and actions that would cause the insulation margin prediction value to fall below the insulation margin threshold after execution are eliminated. For the retained candidate actions, actions that can reduce the load range of the target environmental gas ring main unit are given priority; if there are still multiple actions, actions that reduce the node voltage deviation after execution are further retained; if there are still multiple actions, actions that reduce the line loss after execution are also retained; when necessary, the retained candidate control actions can be combined in sequence to form control action combinations, and combination deduction is performed to select control action combinations that simultaneously satisfy the following conditions: the already powered load is kept powered, the load range of the target environmental gas ring main unit is not expanded, the insulation margin prediction value is not lower than the insulation margin threshold, the node voltage deviation is reduced, and the line loss is reduced. Finally, the control action combinations are converted into power supply stability adjustment command outputs according to the action execution order.

[0045] For example, a target environmental gas ring main unit currently supplies power to 12 load access points on two branches. The system predicts that its insulation margin will approach the threshold in the next 20 minutes. The system lists four candidate actions: closing the tie switch to transfer some load, opening the sectionalizing switch to isolate the terminal branch, switching on reactive power compensation equipment to improve voltage quality, and partial load switching. Analysis reveals that while closing the tie switch can reduce line losses, it will cause another load to flow back into the target ring main unit's load range, so it is eliminated. The sectionalizing switch action will cause some already supplied loads to lose power, so it is also eliminated. Of the two remaining actions, load switching can significantly reduce the target ring main unit's load range and reduce node voltage deviation, while reactive power compensation switching can improve voltage but cannot reduce the target ring main unit's load range. Therefore, the system prioritizes the former and converts it into a power supply stability adjustment command.

[0046] like Figure 1 and Figure 5 As shown, the active defense method includes: checking the predicted insulation margin values ​​at each moment in chronological order; determining the moment when the predicted insulation margin value first falls below the insulation margin threshold to obtain the target moment; defining the target moment and subsequent periods below the insulation margin threshold as risk periods; extracting the current state sequence before the start of the risk period, as well as the real-time gas insulation state parameters and real-time electrical operation parameters within the risk period to form pre-risk gas state records, pre-risk operation state records, gas state records during the risk period, and operation state records during the risk period; and comparing the pre-risk gas state records with the gas state records during the risk period to determine whether the directions of change in moisture content, gas pressure, and temperature are correct. Changes occur; compare the pre-risk operating status record with the in-risk operating status record to determine whether the direction of load current change, voltage fluctuation direction, and sequence of switching actions have changed. If at least one of the following changes—moisture content change direction, gas pressure change direction, and temperature change direction—has changed, and the load current change direction, voltage fluctuation direction, and sequence of switching actions have not changed, the risk source is determined to be a change in gas insulation condition. If the load current change direction or voltage fluctuation direction has changed, and the moisture content change direction, gas pressure change direction, and temperature change direction have not changed, the risk source is determined to be a change in operating load. If the sequence of switching actions has changed, and... When changes occur before the start of a risk period or during the duration of a risk period, the risk source is identified as a change in switch operation. If two or three of the above risk sources are simultaneously present, an active defense command is generated in the following order: first, reduce the load on the target environmental gas ring main unit; then, stop all switch operations except those necessary to maintain the current power supply path connection; and finally, implement load switching. When the risk source is a change in gas insulation status, an active defense command is generated to reduce the load on the target environmental gas ring main unit and prohibit the transfer of new loads. When the risk source is a change in operating load, a command is generated to switch the load whose load current change direction has been altered from the target environmental gas ring main unit to an adjacent power supply path. The active defense command is generated when the risk source is a change in switch operation. It stops all switch operations except those necessary to maintain the current power supply path connection and keeps the current power supply path unchanged. After executing the active defense command, the insulation margin prediction value for the subsequent period is re-acquired. When the re-acquired insulation margin prediction value is no longer lower than the insulation margin threshold, the current operating state is maintained. When the re-acquired insulation margin prediction value is still lower than the insulation margin threshold, the active defense command to reduce the load on the target environmental gas ring main unit is continued to be executed. When the re-acquired insulation margin prediction value continues to decrease, the active defense command to make the target environmental gas ring main unit withdraw from load transfer or withdraw from operation is executed.

[0047] It is important to note that the proactive defense method is used to analyze the source of risk and implement tiered measures when the predicted insulation margin has reached or is about to reach the risk threshold. First, the predicted insulation margin values ​​for each time period within the future timeframe are checked chronologically to determine the moment when the predicted insulation margin first falls below the insulation margin threshold. This moment and subsequent periods continuously falling below the threshold are defined as risk periods. Next, the current state sequence before the start of the risk period, as well as the real-time gas insulation state parameters and real-time electrical operating parameters during the risk period, are extracted to form pre-risk gas state records, pre-risk operating state records, gas state records during the risk period, and operating state records during the risk period. By comparing the records before and after the risk, it is determined whether the direction of change in moisture content, gas pressure, and temperature has changed, and whether the direction of change in load current, voltage fluctuation, and the sequence of switching actions has changed. This distinguishes whether the risk primarily originates from changes in gas insulation state, operating load, switching actions, or a combination of factors. When the risk source is determined to be a change in gas insulation state, it indicates a more significant deterioration in the insulation state of the equipment itself. In this case, priority is given to reducing the target loop. The system employs defensive measures to ensure the gas-fired ring main unit (RMN) bears the load and prohibits the transfer of new loads. When the risk source is determined to be a change in operating load, it indicates that the risk is more likely triggered by changes in external operating load. In this case, the load whose load current change direction has changed can be switched to an adjacent power supply path. When the risk source is determined to be a change in switch operation, it indicates that frequent or abnormal switch operations may induce insulation risks. In this case, all switch operations except those necessary to maintain the connection of the current power supply path should be stopped, and the current power supply path should remain unchanged. If two or three of the above risk sources are simultaneously present, an active defense command is generated in the following order: first, reduce the load borne by the target environmental gas RMN; then, stop unnecessary switch operations; and finally, implement load switching to prioritize reducing equipment pressure and operational disturbances. After executing the active defense command, the insulation margin prediction value for subsequent periods is re-acquired. If the insulation margin is no longer lower than the preset threshold after re-prediction, the current operational status after defense is maintained. If the re-prediction is still lower than the threshold, load reduction measures are continued. If the re-prediction shows that the insulation margin continues to decrease, a defense command is further executed to remove the target environmental gas RMN from load transfer or to take it out of operation to prevent the insulation fault from escalating.

[0048] For example, the system predicts that the insulation margin of a target environmental gas ring main unit will first fall below the threshold at minute 8 within the next 15 minutes, and continue to decline for the following 5 minutes. Therefore, minutes 8 to 13 are identified as the risk period. After comparing the state records before and after the risk, the system finds that the direction of change in moisture content has changed from basically stable to continuously increasing, and the direction of change in gas pressure has changed from a slow decline to an accelerated decline. However, the direction of change in load current and voltage fluctuation remains basically unchanged, and the sequence of switching actions has also remained unchanged. Therefore, the source of risk is determined to be a change in the gas insulation state. Based on this, the system generates an active defense command to reduce the load on the target ring main unit and prohibit the transfer of new loads. If the insulation margin continues to decline after execution and a new prediction is made, the load reduction continues. If the prediction continues to worsen, the system commands the target ring main unit to withdraw from load transfer or to stop operation, so as to control the insulation risk within a safe range. Example 2:

[0049] Taking a 10kV dual-power ring network supply area as an example, four identical environmentally friendly gas ring network cabinets are configured in the area, with the target equipment being ring network cabinet No. 3. The system continuously collects historical information for 12 months at a 10-minute cycle, obtaining a total of 210,240 sets of data. The data includes load current, node voltage, switch operation, moisture content, gas pressure, temperature, and insulation margin calculated accordingly. Trend analysis is performed on the historical insulation margin sequence, assuming a normal fluctuation range of ±1.5%, identifying 36 degradation process segments where the insulation margin continuously decreases and the cumulative decrease exceeds 4%. Based on the average insulation margin decrease rate of 0.38% / h in these process segments and the gas state parameter fluctuation inertia time of approximately 50 minutes, an adaptive window length of 2 hours is set, and 128 historical degradation segments are formed after slicing. The insulation margin sequence 2 hours after the end of each segment is extracted as subsequent insulation evolution data, establishing an insulation degradation historical sample library. During the real-time operation phase, the real-time status data of the target No. 3 ring main unit for the most recent 2 hours was acquired. After time alignment, missing value repair, and noise reduction, the following results were obtained: the moisture content increased from 148ppm to 172ppm, the air pressure decreased from 0.142MPa to 0.138MPa, the temperature increased from 31.6°C to 33.1°C, and the insulation margin decreased from 23.4% to 20.8%. During the same period, the load current increased from 246A to 268A, and the bus voltage fluctuation range was approximately ±2.1%. After extracting the change rate, the moisture change rate was obtained as +2.0ppm / 10min, the air pressure change rate as -0.00033MPa / 10min, the temperature change rate as +0.13°C / 10min, and the original insulation margin change rate as -0.22% / 10min. After eliminating the operating condition disturbance by combining the load current and voltage fluctuations, the insulation margin change rate after eliminating the operating condition disturbance was obtained as -0.18% / 10min. Based on this, the current status sequence was constructed. The current state sequence was matched with 128 historical degradation segments in the sample library for similarity degradation. The segments were then filtered based on the order of change direction, magnitude, duration, and the relative occurrence of the insulation margin change rate, retaining three target degradation segments with cumulative amplitude differences of 1.84, 2.11, and 2.36, respectively. Since the load current change direction sequence, voltage fluctuation direction sequence, and switching action sequence corresponding to these three segments are consistent with the current state, their corresponding subsequent insulation evolution data were directly used for margin estimation. Using the current actual insulation margin of 20.8% as the prediction starting point, the predicted insulation margin values ​​for every 30 minutes within the next 2 hours were obtained as 19.9%, 19.1%, 18.2%, and 17.3%, respectively. Assuming an insulation margin threshold of 18.5%, the predicted value first falls below the threshold after 90 minutes, indicating an insulation risk for the target ring main unit. Based on the above predictions, coordinated control is implemented. Currently, ring main unit No. 3 carries loads from sections L4, L5, and L6, totaling 382kW. The maximum voltage deviation at the original nodes in the area is 6.2%, and the line loss is 18.6kW. After extrapolating the candidate actions, we obtain the following: Action A is to transfer 96kW of load from L6. After execution, the target unit load will decrease to 286kW, the voltage deviation will decrease to 4.8%, the line loss will decrease to 16.9kW, and the minimum predicted insulation margin will be 18.7%. Action B is to transfer a total of 168kW of loads from L5 and L6. After execution, the target unit load will decrease to 214kW, the voltage deviation will decrease to 4.1%, the line loss will decrease to 15.8kW, and the minimum predicted insulation margin will be 19.1%. Action C is to put 200kW of load into operation. After the var reactive power compensation device was implemented, the voltage deviation decreased to 5.0%, the line loss decreased to 17.2kW, and the minimum predicted insulation margin was 18.6%. Considering the conditions of "no expansion of load range, predicted insulation margin not lower than the threshold, reduction of node voltage deviation, and reduction of line loss," the combined control scheme of action B + action C was finally selected, generating the adjustment command: first close the tie switch, then disconnect the original power supply section switch, transfer L5 and L6 to the adjacent power supply path, and then activate the 200kvar reactive power compensation device. After implementation, the load on the target cabinet decreased to 214kW, the node voltage deviation further decreased to 3.3%, the line loss decreased to 14.7kW, and the minimum predicted insulation margin increased to 19.2%. In subsequent monitoring, 50 minutes after the above adjustments were implemented, the system predicted again that the insulation margin for the next 2 hours would be 18.8%, 18.2%, 17.5%, and 16.8%, still below the threshold after 60 minutes. Comparing the state before and after the risk period, it was found that the direction of change in moisture content changed from basically stable to continuously increasing, and the direction of change in gas pressure changed from stable to decreasing. However, the direction of change in load current, voltage fluctuation, and the sequence of switching actions remained unchanged. Therefore, the source of risk was determined to be a change in the gas insulation state. Based on this, the system generated an active defense command: prohibiting the transfer of new load to the target ring main unit and further switching 72kW of the L4 load to an adjacent power supply path. After execution, the load of the target unit dropped to 142kW, and the insulation margin for the next 2 hours was re-predicted to be 19.2%, 18.9%, 18.7%, and 18.6%, all of which were not lower than the threshold. The target ring main unit entered a state of controlled insulation risk operation.

[0050] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended embodiments and their equivalents.

Claims

1. An intelligent control method for an environmentally friendly gas ring main unit used for power supply stability regulation, the method comprising: Historical information acquisition: Obtain historical information by acquiring electrical operating parameters, gas insulation status parameters, and insulation margins from the past operation of the environmental protection gas ring main unit; The method is characterized in that it further includes: Sample library establishment: Based on historical information, historical degradation fragments in the insulation degradation process are extracted through sample construction methods, and an insulation degradation historical sample library is established to store historical degradation fragments and subsequent insulation evolution data, and to establish the correspondence between historical degradation fragments and subsequent insulation evolution data; Real-time status acquisition: Determine the target environmentally friendly gas ring main unit, and acquire the real-time gas insulation status parameters and real-time electrical operation parameters of the target environmentally friendly gas ring main unit. The target environmentally friendly gas ring main unit corresponds to the environmentally friendly gas ring main unit that generated historical information. The real-time gas insulation status parameters are converted into a state sequence that can characterize the current insulation change trend through the state characterization method to obtain the current state sequence. Insulation trend prediction: Preset future time period, based on the current state sequence and the insulation degradation history sample library, the trend prediction method is used to perform similar degradation matching on the current state sequence and determine the historical degradation segment to obtain the target degradation segment. Based on the target degradation segment and the insulation degradation history sample library and corresponding relationship, the margin extrapolation method is used to calculate the insulation margin prediction value in the future time period to obtain the prediction result. Adjustment command generation: Based on real-time electrical operating parameters and prediction results, a power supply stability adjustment command is generated through a collaborative control method. The collaborative control method takes voltage deviation and network loss as optimization targets and insulation margin prediction value as a safety constraint. Active defense execution: Real-time monitoring of prediction results. When the predicted insulation margin value is detected to be lower than the preset insulation margin threshold, active defense instructions are generated and executed through active defense methods to bring the target environmental gas ring main unit into a state of controlled insulation risk operation.

2. The intelligent control method for environmentally friendly gas ring main unit for power supply stability regulation according to claim 1, characterized in that: The sample construction method includes: extracting a sequence of insulation margin changes over time based on historical information to obtain a historical insulation margin sequence; performing trend analysis based on the historical insulation margin sequence to identify various time segments in the historical insulation margin sequence where the insulation margin continuously decreases and the decrease exceeds the fluctuation range of normal equipment operation to obtain historical degradation process segments; slicing the historical degradation process segments using an adaptive time window to obtain several windows and the trace moisture content sequence, air pressure sequence, and temperature sequence within the windows; each window and the trace moisture content sequence, air pressure sequence, and temperature sequence within the window constitute a single historical degradation segment; extracting a historical insulation margin sequence of a fixed duration after the end time of the historical degradation segment to obtain subsequent insulation evolution data corresponding to the historical degradation segment; and establishing a historical insulation degradation sample to store all historical degradation segments and their corresponding subsequent insulation evolution data.

3. The intelligent control method for environmentally friendly gas ring main unit for power supply stability regulation according to claim 2, characterized in that: The length of the adaptive time window is determined based on the rate of decrease in insulation margin and the fluctuation inertia time of gas state parameters during the historical degradation process, and the fixed duration corresponds to the length of the adaptive time window.

4. The intelligent control method for environmentally friendly gas ring main unit for power supply stability regulation according to claim 1, characterized in that: The state characterization method includes: preprocessing the real-time gas insulation state parameters and real-time electrical operation parameters to obtain real-time state basic data for time alignment, missing value repair, and noise reduction; extracting the rate of change of micro-moisture content, gas pressure, temperature, and insulation margin based on the real-time state basic data; eliminating the influence of operating conditions on the rate of change of insulation margin by combining the load current, voltage fluctuation, and switching action information in the real-time electrical operation parameters to obtain the de-operating condition disturbance state component; arranging the de-operating condition disturbance state components at multiple consecutive time points in a temporal sequence based on the sampling step size; constructing a multi-dimensional state vector that can characterize the insulation change trend based on the parameter change direction, change amplitude, and duration at each time point; and combining the multi-dimensional state vectors at multiple consecutive time points in chronological order to obtain the current state sequence for subsequent similarity degradation matching.

5. The intelligent control method for environmentally friendly gas ring main unit for power supply stability regulation according to claim 1, characterized in that: The trend prediction method includes: aligning the current state sequence with each historical degradation segment in the insulation degradation historical sample library according to the same parameter order and the same time order; comparing the direction of change, magnitude of change, order of occurrence of continuous change time periods, and primacy relationship of corresponding parameters in the current state sequence and each historical degradation segment at corresponding times; eliminating historical degradation segments where corresponding parameters have opposite directions of change at the same corresponding time, different order of occurrence of continuous change time periods, or different primacy relationships; based on the historical degradation segments retained after elimination, accumulating the difference in magnitude of change of each corresponding parameter at each corresponding time in chronological order, and retaining the historical degradation segment with the smallest accumulation result to obtain one or more target degradation segments; when multiple target degradation segments are obtained, using all of the multiple target degradation segments as input for subsequent margin extrapolation; the parameter order includes the rate of change of micro-moisture content, the rate of change of air pressure, the rate of change of temperature, and the rate of change of insulation margin.

6. The intelligent control method for environmentally friendly gas ring main unit for power supply stability regulation according to claim 1, characterized in that: The margin extrapolation method includes: retrieving subsequent insulation evolution data corresponding to the target degradation segment from a historical insulation degradation sample library, extracting the subsequent insulation margin sequence arranged in chronological order, determining the actual insulation margin of the target environmental gas ring main unit corresponding to the end time of the current state sequence as the predicted starting insulation margin, calculating the insulation margin change of each subsequent time in the corresponding subsequent insulation margin sequence relative to the starting time when there is only one target degradation segment, and applying it sequentially to the predicted starting insulation margin to obtain the initial insulation margin prediction value, when there are multiple target degradation segments, calculating the insulation margin change of each subsequent time in the subsequent insulation margin sequence corresponding to each target degradation segment relative to its respective starting time, selecting the change in the middle position of multiple changes at the same time position as the representative change, and... The initial insulation margin prediction value is obtained by sequentially applying the predicted starting insulation margin in chronological order. The load current change direction sequence, voltage fluctuation direction sequence, and switch action sequence within the time period corresponding to the current state sequence of the target environmental gas ring main unit are extracted and compared with the corresponding sequences within the historical time period corresponding to each target degradation segment. When the comparison results of each sequence are the same, the initial insulation margin prediction value is used as the insulation margin prediction value. When the comparison results of each sequence are different, the target degradation segment that is inconsistent with the current state sequence is removed, and the insulation margin prediction value is recalculated based on the retained target degradation segment. If no target degradation segment is retained after removal, the initial insulation margin prediction value obtained before removal is retained, and the insulation margin prediction value is obtained by extracting the prediction values ​​of each moment corresponding to the future time period based on the time series of the initial insulation margin prediction value.

7. The intelligent control method for environmentally friendly gas ring main unit for power supply stability regulation according to claim 1, characterized in that: The collaborative control method includes: acquiring the on / off status of each switch, the on / off relationship of each line, the node voltage of each node, the line current of each line, the load access relationship of each load access point, and the activation status of each reactive power compensation device within the power supply area where the target environmental gas ring network cabinet is located; determining the current power supply path based on the on / off status of each switch and the on / off relationship of each line; determining the load range supplied by the target environmental gas ring network cabinet based on the load access relationship of each load access point; acquiring candidate control actions that can change the current power supply path or load range, including sectional switch opening, sectional switch closing, tie switch opening, tie switch closing, load switching, and reactive power compensation device switching; performing a deduction on each candidate control action to obtain the power supply path, load range, node voltage, line current, and line loss after execution; and eliminating those that lead to changes after execution. Candidate control actions that cause power loss to the previously supplied loads, expand the load range of the target environmental gas ring main unit, or cause the insulation margin prediction value to fall below the insulation margin threshold are selected. Based on the reserved candidate control actions, priority is given to retaining candidate control actions that reduce the load range of the target environmental gas ring main unit. When there are multiple candidate control actions, the candidate control action that reduces the node voltage deviation after execution is retained. When there are still multiple candidate control actions after retention, the candidate control action that reduces the line loss after execution is retained. The retained candidate control actions are combined in sequence and combined deduction is performed to select the combination of control actions that simultaneously satisfy the following conditions: the previously supplied loads are kept powered, the load range of the target environmental gas ring main unit is not expanded, the insulation margin prediction value is not lower than the insulation margin threshold, the node voltage deviation is reduced, and the line loss is reduced. This combination of control actions is then converted into power supply stabilization adjustment commands according to the action execution order.

8. The intelligent control method for environmentally friendly gas ring main unit for power supply stability regulation according to claim 7, characterized in that: The active defense method includes: checking the predicted insulation margin values ​​at each moment in chronological order; determining the moment when the predicted insulation margin value first falls below the insulation margin threshold to obtain the target moment; defining the target moment and subsequent periods below the insulation margin threshold as risk periods; extracting the current state sequence before the start of the risk period, as well as the real-time gas insulation state parameters and real-time electrical operation parameters within the risk period to form pre-risk gas state records, pre-risk operation state records, gas state records during the risk period, and operation state records during the risk period; and comparing the pre-risk gas state records with the gas state records during the risk period to determine whether changes in the direction of moisture content change, gas pressure change, and temperature change have occurred. Changes occur; compare the pre-risk operating status record with the in-risk operating status record to determine whether the direction of load current change, voltage fluctuation direction, and sequence of switching actions have changed. If at least one of the following changes—moisture content change direction, gas pressure change direction, and temperature change direction—has changed, and the load current change direction, voltage fluctuation direction, and sequence of switching actions have not changed, the risk source is determined to be a change in gas insulation condition. If the load current change direction or voltage fluctuation direction has changed, and the moisture content change direction, gas pressure change direction, and temperature change direction have not changed, the risk source is determined to be a change in operating load. If the sequence of switching actions has changed, and the change is not specified, the risk source is determined to be a change in operating load. When a change occurs before the start of a risk period or during the duration of a risk period, the risk source is determined to be a change in switch operation. When two or three of the above risk sources are simultaneously present, an active defense command is generated in the following order: first, reduce the load on the target environmental gas ring main unit; then, stop all switch operations except those necessary to maintain the current power supply path connection; and finally, implement load switching. When the risk source is a change in gas insulation status, an active defense command is generated to reduce the load on the target environmental gas ring main unit and prohibit the transfer of new loads. When the risk source is a change in operating load, an active defense command is generated to switch the load whose load current change direction has changed from the target environmental gas ring main unit to the adjacent power supply path. Active defense command: When the risk source is a change in switch operation, generate an active defense command to stop all switch operations except those necessary to maintain the current power supply path connection and keep the current power supply path unchanged. After executing the active defense command, reacquire the insulation margin prediction value for the subsequent time period. When the reacquired insulation margin prediction value is no longer lower than the insulation margin threshold, maintain the current operating state. When the reacquired insulation margin prediction value is still lower than the insulation margin threshold, continue to execute the active defense command to reduce the load borne by the target environmental gas ring main unit. When the reacquired insulation margin prediction value continues to decrease, execute the active defense command to make the target environmental gas ring main unit withdraw from load transfer or withdraw from operation.

Images