Voltage out-of-limit identification method and device for low-voltage transformer area

Abstract

The invention discloses a voltage out-of-limit identification method and device for a low-voltage transformer area, and the method comprises the steps: recognizing whether voltage out-of-limit exists or not according to collected user voltage data, and recognizing a photovoltaic scene based on collected gateway data if the voltage out-of-limit exists; carrying out distribution transformer load rate analysis, distribution transformer three-phase imbalance analysis, distribution transformer power factor analysis, distribution transformer gear analysis and transformer area power supply radius analysis on the basis of the gateway data to obtain a load rate analysis result, a three-phase imbalance analysis result, a power factor analysis result, a gear analysis result and a power supply radius analysis result; a decision tree model is used for carrying out multi-factor correlation analysis on multiple analysis results to obtain the contribution ratio of each factor to voltage out-of-limit, and gateway data is combined with the decision tree model to realize comprehensive study and judgment and accurate attribution of multiple factors, so that the accuracy and comprehensiveness of voltage out-of-limit cause identification are improved, and the accuracy and comprehensiveness of voltage out-of-limit cause identification are improved. And a solid support is provided for formulating differentiated and effective treatment strategies.

Description

Technical Field

[0001] This invention relates to the field of voltage over-limit identification technology, and in particular to a method and apparatus for identifying voltage over-limit in low-voltage distribution areas. Background Technology

[0002] Distributed photovoltaic (PV) power generation, with its advantages of localized consumption, small footprint, and simple control methods, has been integrated into low-voltage distribution networks in a high-density, large-scale manner. In today's complex and ever-changing energy landscape, distributed PV power generation has been widely and deeply applied and promoted due to its significant advantages of being clean, renewable, efficient, and flexible. The connection of a large number of distributed PV devices, especially on the low-voltage side of distribution network areas, is showing a rapid expansion trend, promoting the popularization of new energy sources and optimizing the energy structure. However, the accompanying technical and operational challenges are also becoming increasingly prominent, particularly manifested in problems such as severe voltage exceeding limits in distribution areas and significant degradation of power quality, placing enormous pressure on the stable operation of the distribution network.

[0003] With the grid connection of numerous low-voltage distributed photovoltaic (PV) systems, the original voltage balance of the distribution network has been disrupted. Under sufficient sunlight, the output power of PV devices increases significantly. If this exceeds the load absorption capacity of the distribution area, the excess energy will be fed back to the grid, causing a sharp rise in the distribution area's outlet voltage (the low-voltage output voltage of the distribution area), frequently exceeding the standard allowable voltage fluctuation range. According to relevant standards, the allowable deviation of the distribution area's outlet voltage is ±7% of the rated voltage. However, in actual operation, voltage exceeding limits due to distributed PV access occurs frequently, with voltage deviations in some high PV penetration areas even exceeding 10%. This not only affects the safe and stable operation of power equipment but also severely weakens the power quality experience for users. For example, in a certain area of ​​Fujian, users connected to distributed PV systems typically disable reactive power regulation and set the output voltage to 5%-10% of the grid voltage to ensure they can generate power. This leads to voltage exceeding limits for surrounding users, resulting in damage to their equipment. Furthermore, the intermittent and fluctuating nature of PV power generation causes output power to fluctuate rapidly with changes in sunlight intensity and weather conditions, resulting in frequent voltage fluctuations and voltage flicker, further exacerbating power quality problems.

[0004] The causes of voltage exceeding limits in transformer substations are complex and varied, and cannot be explained by a single factor. Excessively thin power lines increase line resistance, leading to greater voltage loss when transmitting the same power, resulting in lower voltage at the substation's end. Simultaneously, photovoltaic (PV) power injection can easily cause excessive voltage. For traditionally single-source-powered substations, to ensure normal supply voltage at the line's end, transformer tap settings typically result in excessively high voltage at the substation's cut-off point. This transformer tap setting lacks a reasonable and flexible adjustment mechanism, making it difficult to adapt to load changes and the dynamic characteristics of PV integration, further exacerbating voltage instability. Under high load conditions, the increased load on transformers and lines leads to greater voltage drop, which is detrimental to substation voltage stability. Furthermore, improper capacitor compensation configurations may result in excess or insufficient reactive power, affecting the robust control of voltage levels. More critically, the integration of distributed PV directly alters the power flow distribution in the substation; if the capacity and connection location are not properly selected, voltage exceeding limits and power quality problems can easily occur.

[0005] Current criteria for determining voltage exceedances in transformer substations are significantly inadequate. Traditional criteria are simplistic, generally employing the single logic that "if there is photovoltaic (PV) connection and the voltage exceedance is due to PV," failing to comprehensively consider the complex interplay of various influencing factors. In reality, in many cases of voltage exceedances, PV is not the sole or dominant cause. This misattribution leads to insufficiently targeted remedial measures, wasting significant human, material, and financial resources, and failing to fundamentally resolve the issues of voltage anomalies and power quality degradation. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to provide a method and device for identifying voltage over-limit in low-voltage distribution areas, which can improve the accuracy and comprehensiveness of identifying the causes of voltage over-limit.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A method for assessing the operational status of a photovoltaic power generation system, comprising the following steps: Obtain the evaluation indicators for the photovoltaic power generation system to be evaluated; The evaluation indicators were subjectively weighted using a subjective weighting method to obtain subjective weights; The evaluation indicators are objectively weighted using multiple objective weighting methods to obtain objective weights; The subjective weights and the objective weights are combined and weighted to obtain the combined weights; The evaluation index and the combined weights are used to form a feature vector, which is then input into the trained support vector machine model to evaluate the operating status, thereby obtaining the operating status evaluation result of the photovoltaic power generation system to be evaluated.

[0008] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is as follows: A low-voltage distribution area voltage over-limit identification device, comprising: The data acquisition module is used to collect user voltage data and gate data in the low-voltage distribution area; The voltage limit exceedance detection module is used to identify whether there is a voltage limit exceedance based on the user voltage data and obtain the detection result; A photovoltaic scene identification module is used to identify a photovoltaic scene based on the threshold data if the identification result indicates that there is a voltage over-limit. The factor analysis module is used to perform distribution transformer load rate analysis, distribution transformer three-phase imbalance analysis, distribution transformer power factor analysis, distribution transformer tap analysis, and distribution area power supply radius analysis based on the aforementioned data, and to obtain load rate analysis results, three-phase imbalance analysis results, power factor analysis results, tap analysis results, and power supply radius analysis results. The multi-factor correlation analysis module is used to perform multi-factor correlation analysis on the photovoltaic scenario, the load rate analysis results, the three-phase imbalance analysis results, the power factor analysis results, the tap analysis results, and the power supply radius analysis results using a decision tree model, so as to obtain the contribution ratio of each factor to voltage over-limit.

[0009] The beneficial effects of this invention are as follows: Based on the collected user voltage data, it identifies whether voltage limits are exceeded, and obtains the identification result. If voltage limits are exceeded, it identifies the photovoltaic scenario based on the collected threshold data, and performs distribution transformer load rate analysis, distribution transformer three-phase imbalance analysis, distribution transformer power factor analysis, distribution transformer tap analysis, and distribution area power supply radius analysis based on the threshold data, obtaining load rate analysis results, three-phase imbalance analysis results, power factor analysis results, tap analysis results, and power supply radius analysis results. A decision tree model is then used to analyze the photovoltaic scenario, load rate analysis results, three-phase imbalance analysis results, and power factor analysis results. Multi-factor correlation analysis was conducted using the results of rate factor analysis, tap level analysis, and power supply radius analysis to obtain the contribution ratio of each factor to voltage over-limit. By combining key data with a decision tree model, a comprehensive judgment and accurate attribution of multiple factors can be achieved when voltage over-limit is identified. Through the hierarchical classification logic of the decision tree, the contribution of each factor to voltage over-limit is quantitatively analyzed, which can clearly identify the dominant cause of voltage over-limit and avoid misattribution to photovoltaics. This improves the accuracy and comprehensiveness of voltage over-limit cause identification and provides solid support for the formulation of differentiated and effective governance strategies. Attached Figure Description

[0010] Figure 1 This is a flowchart of a low-voltage distribution area voltage over-limit identification method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a low-voltage distribution area voltage over-limit identification device according to an embodiment of the present invention; Figure 3This is an equivalent diagram of photovoltaic grid connection in a low-voltage distribution area voltage over-limit identification method according to an embodiment of the present invention; Figure 4 This is a test area topology diagram in a low-voltage distribution area voltage over-limit identification method according to an embodiment of the present invention. Detailed Implementation

[0011] To explain in detail the technical content, objectives, and effects of the present invention, the following description is provided in conjunction with the embodiments and accompanying drawings.

[0012] Before detailing the embodiments of this application, some related concepts will first be explained: Key data: Voltage, current, active power, reactive power, power factor, and other values ​​obtained from sampling and processing of PT (voltage transformer) and CT (current transformer) installed on the low-voltage side of the distribution transformer. The sampling time is 15 minutes per point, with a total of 96 data sampling points per day. These data are uploaded to the data server through the distribution area centralized data acquisition device.

[0013] Voltage exceeding limits: refers to the voltage in a power system exceeding the specified rated value.

[0014] The current shortcomings of photovoltaic voltage criteria in existing technologies are mainly reflected in the simplistic content and crude logic of the criteria, as well as insufficient utilization of limited data mining. The judgment results exhibit problems such as insufficient content, ambiguous logic, and low accuracy. The system lacks understanding of the complex interaction between photovoltaic power generation and electricity load, causing its judgment to often remain at the single dimension of installed capacity, ignoring the dynamic impact of actual photovoltaic power generation. At the same time, the impact of three-phase imbalance and reactive power anomalies, which are simply categorized as "other problems," on user voltage has not been effectively identified. This one-size-fits-all classification method limits in-depth analysis of the causes of voltage exceedances. Consequently, the practical application of the criteria is limited, making it difficult to provide accurate fault identification and effective control basis for grid operation.

[0015] To at least solve the above problems, please refer to Figure 1 This invention provides a method for identifying voltage over-limit in low-voltage distribution areas, comprising the following steps: Collect user voltage data and switch data for low-voltage distribution areas; Based on the user voltage data, identify whether there is a voltage over-limit issue and obtain the identification result; If the identification result indicates that there is a voltage limit violation, then the photovoltaic scenario is identified based on the threshold data; Based on the aforementioned data, we perform transformer load rate analysis, transformer three-phase imbalance analysis, transformer power factor analysis, transformer tap position analysis, and transformer supply radius analysis to obtain the results of load rate analysis, three-phase imbalance analysis, power factor analysis, tap position analysis, and supply radius analysis. A decision tree model was used to perform a multi-factor correlation analysis on the photovoltaic scenario, the load rate analysis results, the three-phase imbalance analysis results, the power factor analysis results, the tap analysis results, and the power supply radius analysis results to obtain the contribution ratio of each factor to voltage exceedance.

[0016] As can be seen from the above description, the beneficial effects of the present invention are as follows: Based on the collected user voltage data, it identifies whether there is a voltage limit violation and obtains the identification result. If a voltage limit violation is identified, it identifies the photovoltaic scenario based on the collected gate data, and performs distribution transformer load rate analysis, distribution transformer three-phase imbalance analysis, distribution transformer power factor analysis, distribution transformer tap analysis, and distribution area power supply radius analysis based on the gate data, obtaining load rate analysis results, three-phase imbalance analysis results, power factor analysis results, tap analysis results, and power supply radius analysis results. A decision tree model is then used to analyze the photovoltaic scenario, load rate analysis results, and three-phase imbalance analysis. The results, power factor analysis, tap position analysis, and power supply radius analysis are used to conduct multi-factor correlation analysis to obtain the contribution ratio of each factor to voltage over-limit. By combining key data with a decision tree model, a comprehensive judgment and accurate attribution of multiple factors can be achieved when voltage over-limit is identified. Through the hierarchical classification logic of the decision tree, the contribution of each factor to voltage over-limit is quantitatively analyzed, which can clearly identify the dominant cause of voltage over-limit, avoid misattributing it to photovoltaics, and thus improve the accuracy and comprehensiveness of voltage over-limit cause identification, providing solid support for the formulation of differentiated and effective governance strategies.

[0017] Furthermore, the gate data includes the gate three-phase voltage value and the gate three-phase current value; Identifying photovoltaic scenarios based on the aforementioned gateway data includes: If the three-phase current values ​​at the threshold are reversed, the photovoltaic scenario is identified as one where the photovoltaic power generation is greater than the power consumption of the distribution area. If they are not reversed, the historical data of the three-phase voltage values ​​at the threshold during the same period of the preset time period are compared to determine whether the three-phase voltage values ​​at the threshold have risen abnormally. If they have, the photovoltaic scenario is identified as one where the power consumption of the distribution area is greater than the photovoltaic power generation and is greater than zero. If they have not, the photovoltaic scenario is identified as one where the power consumption of the distribution area is greater than the photovoltaic power generation and is equal to zero.

[0018] As described above, the impact of photovoltaic (PV) distribution areas is not only on the combined effects of user power levels and PV capacity, but also on the distribution area topology and PV connection points. At the distribution area junction, this is reflected in the junction current and voltage characteristics. Due to the lack of direct power data, the relationship between PV output and power consumption is indirectly determined by analyzing the junction current direction and voltage-current correlation characteristics, combined with voltage variation patterns in typical scenarios, thus achieving accurate PV scenario identification.

[0019] Furthermore, the threshold data includes the daily maximum load rate of the distribution transformer, the daily average load rate of the distribution transformer, and the distribution transformer capacity; Based on the aforementioned data, a distribution transformer load rate analysis was performed, yielding the following results: If the maximum daily load rate of the distribution transformer exceeds the first preset percentage, the load rate analysis result is identified as a risk of distribution transformer overload. Obtain the number of households in the transformer substation area, and calculate the average capacity per household based on the number of households in the transformer substation area and the transformer capacity; If the average capacity per household is less than the preset capacity, and the average daily load rate of the distribution transformer exceeds the second preset percentage, then the load rate analysis result is identified as low average capacity per household exacerbating the voltage over-limit problem.

[0020] As can be seen from the above description, the impact of load rate on voltage over-limit issues can be effectively clarified by using the daily maximum load rate of the distribution transformer, the average capacity per household, and the daily average load rate of the distribution transformer.

[0021] Furthermore, the threshold data also includes the daily maximum three-phase imbalance. Based on the aforementioned threshold data, a three-phase imbalance analysis of the distribution transformer was performed, yielding the following results: Obtain the transformer wiring type; If the transformer wiring type is Yyn0 wiring, and the maximum daily three-phase imbalance exceeds the third preset percentage, and the transformer daily average load rate exceeds the second preset percentage, then the three-phase imbalance analysis result is identified as a phase voltage deviation problem caused by three-phase imbalance. If the transformer wiring type is Dyn11 wiring, and the maximum daily three-phase imbalance exceeds the fourth preset percentage, and the transformer's daily average load rate exceeds the second preset percentage, then the three-phase imbalance analysis result is identified as a phase voltage deviation problem caused by three-phase imbalance.

[0022] As described above, different types of power distribution wiring can be identified, and different wiring types have corresponding judgment thresholds. By superimposing the daily average load rate of the distribution transformer, misjudgments can be effectively avoided, and more accurate three-phase imbalance analysis of the distribution transformer can be achieved.

[0023] Furthermore, the threshold data also includes the distribution transformer power factor; Based on the aforementioned data, a power factor analysis of the distribution transformer was performed, yielding the following results: If the power factor of the distribution transformer is lower than the preset power factor, and the daily average load rate of the distribution transformer exceeds the second preset percentage, then the power factor analysis result is identified as a low power factor of the distribution transformer.

[0024] As can be seen from the above description, since there is no direct user power data, combined with the power factor data of the distribution point, and considering the load characteristics of the distribution area, such as many single-phase or three-phase household workshops connected, their inductive load is large and they lack reactive power compensation, the daily average load rate of the distribution transformer can be used to determine whether the power factor of the distribution transformer is too low.

[0025] Furthermore, based on the aforementioned checkpoint data, a gear shift analysis is performed on the distribution transformer, yielding the following results: Extract one week's worth of gate voltage data from the three-phase voltage values ​​at the gate; Based on the voltage data at the control points, the duration of severe overvoltage and undervoltage is superimposed to clarify the distribution of voltage over-limit issues before gear adjustment; After lowering the transformer speed, ensure that there are no new low-voltage users, and use the original number of overvoltage users as a base. If the number of overvoltage users decreases by more than the fifth preset percentage, the transformer speed adjustment is deemed effective. After adjusting the transformer speed, the number of users with severe overvoltage did not increase significantly. Based on the original number of low-voltage users, if the number of low-voltage users needs to be reduced by more than the sixth preset percentage, then the transformer speed adjustment is deemed reasonable.

[0026] As described above, by using "periodic screening + limit-crossing correlation + gear adjustment verification", the compatibility between the gear and voltage quality can be quickly determined, and voltage limit-crossing problems caused by unreasonable gears can be accurately identified.

[0027] Furthermore, the gateway data also includes GIS data on transformer area topology and user distribution; Based on the aforementioned gateway data, the power supply radius analysis of the transformer substation is performed, and the results of the power supply radius analysis include: Calculate the power supply radius of the transformer substation; Based on the aforementioned transformer area topology and user distribution GIS data, the power supply area type and load density characteristics are determined. A power supply radius threshold is matched based on the power supply area type and the load density characteristics; The power supply radius of the transformer area is compared with the power supply radius threshold to obtain the power supply radius analysis results.

[0028] As described above, matching the power supply radius threshold based on the power supply area type and load density characteristics, and then comparing the power supply radius of the transformer area with the power supply radius threshold, yields the power supply radius analysis results. This approach better reflects the characteristics of different power supply areas and ensures the reliability of the power supply radius analysis.

[0029] Furthermore, a decision tree model is used to perform a multi-factor correlation analysis on the photovoltaic scenario, the load factor analysis results, the three-phase imbalance analysis results, the power factor analysis results, the tap level analysis results, and the power supply radius analysis results, to obtain the contribution percentage of each factor to voltage exceedance, including: The influence of the photovoltaic scenario, the load rate analysis results, the three-phase imbalance analysis results, the power factor analysis results, the tap analysis results, and the power supply radius analysis results is prioritized. Determine the criterion confidence levels for the photovoltaic scenario, the load rate analysis results, the three-phase imbalance analysis results, the power factor analysis results, the tap analysis results, and the power supply radius analysis results; Based on the priority of the degree of impact and the confidence level of the criterion, the photovoltaic scenario, the load rate analysis results, the three-phase imbalance analysis results, the power factor analysis results, the tap analysis results, and the power supply radius analysis results are sorted to obtain a list of possible causes of the transformer area voltage exceeding the limit problem; Input the list of possible causes of the voltage over-limit problem in the transformer area into the decision tree model for multi-factor correlation analysis, and output the contribution ratio of each factor to the voltage over-limit problem.

[0030] As described above, after sorting by the two dimensions of "impact priority + criterion confidence", a list of possible causes of voltage over-limit problems in the transformer area is generated. Then, a decision tree model is used to conduct multi-factor correlation analysis. The decision tree model can refine and distinguish the dynamic matching relationship between photovoltaic power and load, avoiding misjudgment or omission caused by using photovoltaic installed capacity as the sole criterion, and making the attribution of voltage over-limit more in line with the actual scenario.

[0031] Furthermore, it also includes: Based on the contribution percentage of each factor to voltage over-limit, a voltage over-limit diagnostic report is generated using knowledge graphs and large language models.

[0032] As described above, using knowledge graphs and large language models to generate voltage over-limit diagnostic reports can provide a clear and intuitive understanding of the causes of voltage over-limit.

[0033] Please refer to Figure 2 Another embodiment of the present invention provides a low-voltage distribution area voltage over-limit identification device, comprising: The data acquisition module is used to collect user voltage data and gate data in the low-voltage distribution area; The voltage limit exceedance detection module is used to identify whether there is a voltage limit exceedance based on the user voltage data and obtain the detection result; A photovoltaic scene identification module is used to identify a photovoltaic scene based on the threshold data if the identification result indicates that there is a voltage over-limit. The factor analysis module is used to perform distribution transformer load rate analysis, distribution transformer three-phase imbalance analysis, distribution transformer power factor analysis, distribution transformer tap analysis, and distribution area power supply radius analysis based on the aforementioned data, and to obtain load rate analysis results, three-phase imbalance analysis results, power factor analysis results, tap analysis results, and power supply radius analysis results. The multi-factor correlation analysis module is used to perform multi-factor correlation analysis on the photovoltaic scenario, the load rate analysis results, the three-phase imbalance analysis results, the power factor analysis results, the tap analysis results, and the power supply radius analysis results using a decision tree model, so as to obtain the contribution ratio of each factor to voltage over-limit.

[0034] The low-voltage distribution area voltage over-limit identification method and device described above are applicable to low-voltage distribution area voltage over-limit identification scenarios. The specific implementation methods are described below: Please refer to Figure 1 One embodiment of the present invention is as follows: A method for identifying voltage over-limit issues in low-voltage distribution areas includes the following steps: S1. Collect user voltage data and gate data for the low-voltage distribution area.

[0035] The threshold data includes the three-phase voltage value, three-phase current value, daily maximum load rate of the transformer, daily average load rate of the transformer, transformer capacity, daily maximum three-phase imbalance, transformer power factor, and GIS data on transformer topology and user distribution. Daily maximum load rate of the transformer = load capacity corresponding to the daily maximum threshold current / transformer capacity (calculated by the ratio of current to transformer rated current, combined with rated capacity); daily average load rate of the transformer = load capacity corresponding to the daily average threshold current / transformer capacity; daily maximum three-phase imbalance = (maximum current - minimum current) / maximum current. 100%.

[0036] Since the system lacks direct power data, subsequent analysis can be conducted based on existing measurements by leveraging the electrical characteristics of voltage and current (such as Ohm's law and the three-phase imbalance law) and combining scenario logic to deduce power-related effects (such as the reverse voltage-current characteristics during photovoltaic backfeed). No additional data collection is required.

[0037] S2. Identify whether there is a voltage over-limit based on the user voltage data, and obtain the identification result.

[0038] Among them, when a voltage limit violation is detected, there are six types of voltage limit violations: general upper limit violation, upper upper limit violation, severe upper limit violation, general lower limit violation, lower lower limit violation, and severe lower limit violation, as shown in Table 1.

[0039] Table 1 Definition of 0.4kV Voltage Quality Indicators

[0040] Specifically, it is determined whether the user voltage data is greater than or equal to 353.4V and less than 406.6V. If so, the identification result is generated as no voltage limit is exceeded; otherwise, the identification result is generated as voltage limit is exceeded.

[0041] The main causes of voltage exceeding limits in distribution transformer areas are as follows: photovoltaic distribution areas, excessively long power supply radius, high transformer load rate, low transformer load rate, three-phase imbalance in transformers, low transformer power factor, unreasonable transformer tap settings, and suspected Yyn0 transformers. The voltage management system of the power grid PMS3.0 provides access to data on voltage and current at relevant terminals, load rate data, three-phase imbalance conditions, and voltage data on the user side.

[0042] S3. If the identification result indicates that there is a voltage exceeding the limit, then the photovoltaic scenario is identified based on the threshold data.

[0043] Among these, photovoltaic (PV) distribution areas are a typical cause of voltage exceedance issues. The basic principle is that PV inverters are typically configured to ensure efficient power generation and profitability, leading to higher output voltages and a voltage rise around PV users. The equivalent circuit for PV grid connection is as follows: Figure 3 As shown, R + jX For the equivalent anode reactance of the line and transformer, P PV + jQ PV The active and reactive power generated by the photovoltaic inverter. P Load + jQ Load Power consumed by local load. P S + jQ S This represents the residual power flowing through the equivalent impedance.

[0044] When the photovoltaic system is not connected to the distribution network, the voltage at the grid connection point is denoted as Distribution network voltage is denoted as Specifically: ; ; When photovoltaic power is connected, it will increase the voltage in its surrounding area and also cause the line voltage to rise, specifically: ; When the load factor of the transformer area is low, its equivalent voltage deviation as follows: ; Therefore, when the change in the active power output of the inverter exceeds a certain value, it will inevitably cause the grid connection point voltage to exceed the limit. For low-voltage distribution networks, its lines... R Often much larger than its X Therefore, in the distribution network R / X A larger ratio means that the voltage at the grid connection point is more sensitive to changes in active power injection, and therefore the voltage quality problems in the distribution area are more obvious.

[0045] Based on the varying relationships between photovoltaic power output and user electricity consumption, the main scenarios can be categorized into the following three types: 1) Photovoltaic power generation > power consumption in the distribution area: At this time, the photovoltaic power generation is difficult for users within the distribution area to absorb, resulting in photovoltaic backfeeding. This causes a reverse voltage drop on the line between the photovoltaic user and the distribution transformer outlet, leading to an increase in voltage for all users in the distribution area. Therefore, the voltage over-limit problem is mainly driven by the active power output of photovoltaic power, and its typical characteristic is the appearance of reverse current at the distribution area threshold. The specific analysis is as follows: Spatiotemporal distribution characteristics: Voltage overshooting mainly occurs during the midday period (10:00-14:00) when photovoltaic power generation is at its peak, and the voltage rise is more pronounced for users closer to the photovoltaic grid connection node, forming a "stepped" or "U-shaped" voltage distribution.

[0046] Current characteristics: A continuous reverse current appears at the transformer substation connection point. The reverse current value is linearly related to the photovoltaic backfeed power (I=P / U, where U is the rated voltage). When the reverse current exceeds 15% of the transformer's rated current, it may trigger the risk of malfunction of the substation-side relay protection device.

[0047] Voltage rise: For every 10kW increase in photovoltaic backfeed power, the average voltage of the main line in the distribution area rises by 2.5-4V, and the voltage at the end user may exceed 242V (exceeding the standard limit of 220V+7%).

[0048] 2) 0 < Photovoltaic power generation < Electricity consumption in the distribution area: At this time, photovoltaic power generation will still have a certain voltage rise effect on users in the distribution area. However, the voltage exceeding the limit problem may not be directly caused by photovoltaic output. It is also necessary to comprehensively consider the users' own electricity consumption behavior. Therefore, further analysis is needed to clarify the dominant factors of users' voltage exceeding the limit. At this time, the load current at the distribution area is positive, but its value is small, and the load rate of the distribution area is low. However, its three-phase imbalance and other characteristics are prominent. The specific analysis is as follows: Voltage fluctuation frequency characteristics: When photovoltaic power output is affected by clouds, the voltage fluctuation frequency is mostly in the range of 0.1-2Hz (such as the power fluctuation period caused by cumulus cloud obstruction, which is about 30-120 seconds). This forms a superposition effect with the voltage dip caused by motor startup (lasting 0.1-1 seconds) at different time scales.

[0049] Impact of inductive loads: When there are a large number of uncompensated inductive loads (such as wind turbines and water pumps) on the user side, the reactive power generated by photovoltaics is difficult to fully compensate, leading to an increase in reactive current in the line and voltage loss. U = Q × X / U ( Q Reactive power X This is due to an increase in line reactance. For example, a certain distribution area has a photovoltaic output of 30kW (including 10kvar reactive power), but the user's inductive reactive power demand reaches 15kvar, resulting in the terminal voltage being 12V lower than the grid connection point.

[0050] Impact of nonlinear loads: When photovoltaic power output is present simultaneously with nonlinear loads (such as inverter air conditioners and welding machines), it can cause a combined disturbance of voltage sag and dip. Actual measurements in a transformer substation of an industrial park showed that when a welding machine was operating, the superimposed photovoltaic power fluctuations caused a voltage dip of up to 15% for 200ms, triggering a communication interruption in the PLC (Programmable Logic Controller) equipment.

[0051] Three-phase voltage deviation quantification: Under light load conditions, for every 10% increase in three-phase load imbalance, the phase voltage deviation increases by 3-5V. In a certain transformer area, phase A accounts for 60% of the load, while phases B and C each account for 20%. Even if the photovoltaic output is evenly distributed across the three phases, the voltage of phase A is still 18V lower than that of phase C, exceeding the 15% limit in the "Permissible Unbalance of Three-Phase Voltage" (GB / T 15543-2008).

[0052] 3) 0 = Photovoltaic power generation < Electricity consumption in the distribution area: At this point, the voltage exceeding the limit for users in the distribution area is only related to their own electricity consumption behavior and characteristics. Their user characteristics are complex and their coupling characteristics are obvious. It is necessary to thoroughly investigate the users' incorrect electricity consumption behavior and at the same time avoid misjudging the photovoltaic system.

[0053] The identification of photovoltaic scenarios based on the gateway data includes: If the three-phase current values ​​at the threshold are reversed, the photovoltaic scenario is identified as one where the photovoltaic power generation is greater than the power consumption of the distribution area. If they are not reversed, the historical data of the three-phase voltage values ​​at the threshold during the same period of the preset time period are compared to determine whether the three-phase voltage values ​​at the threshold have risen abnormally. If they have, the photovoltaic scenario is identified as one where the power consumption of the distribution area is greater than the photovoltaic power generation and is greater than zero. If they have not, the photovoltaic scenario is identified as one where the power consumption of the distribution area is greater than the photovoltaic power generation and is equal to zero.

[0054] In one optional implementation, the preset time period is 10:00-14:00. This period is typically the time of day with high sunlight intensity and high photovoltaic power generation, during which photovoltaic output has a more significant impact on the electrical parameters of the transformer substation. This facilitates the accurate determination of the photovoltaic power generation and the power consumption of the transformer substation by observing characteristics such as abnormal voltage rises, thereby improving the accuracy and relevance of the determination.

[0055] To determine whether the three-phase voltage value at the aforementioned point has risen abnormally, the following steps are taken: Determine whether the voltage fluctuation curve of the three-phase voltage value at the specified point conforms to the photovoltaic "U" curve. Calculate the curve similarity for the three days before and after the point. The curve similarity calculation method is as follows: Combining the robustness of DTW (Dynamic Time Warping) to temporal offsets with the sensitivity of cosine similarity to shape features, a weighted fusion strategy is adopted: Cosine similarity (weight 60%): Primarily determines the shape matching of the "U" curve, focusing on the consistency of voltage curve trends; DTW similarity (weight 40%): Assists in correcting errors caused by time axis offset, adapting to the non-strict synchronous fluctuations in photovoltaic output. A comprehensive similarity is obtained through weighted calculation, ensuring core curve shape matching while tolerating a certain degree of temporal misalignment, thus improving the accuracy of photovoltaic scenario determination. Its pseudocode is as follows: (1) Data preprocessing Determine the original sequences of curves A and B (the lengths can be different, denoted as m and n, with a maximum of 96). Clean outliers (smooth outliers, remove extreme values); (2) DTW similarity calculation Calculate the DTW distance between curves A and B: Construct a (m+1)×(n+1) distance matrix using dynamic programming, and solve for the cumulative distance of the optimal alignment path; distance DTW dis Convert similarity DTW sim : The range is [0,1], and the larger the value, the more similar the two are.

[0056] (3) Cosine similarity calculation Uniform length: Curves A and B are uniformly divided into 96 points of the same length using Lagrange interpolation; Calculate the cosine value of a vector The range is [-1, 1]; Normalization: This is used to convert to [0,1].

[0057] (4) Overall similarity Weighted fusion yields the overall similarity. sim : The weight is 0.5 and can be adjusted as needed.

[0058] (5) Repeat the calculation of the seven-day curve and calculate the offset voltage at its highest point. Calculate the seven-day curve similarity, remove values ​​with excessive similarity deviation (remove noise caused by factors such as power outages), and calculate the average similarity. Using the midpoint value of 380 in the voltage over-limit standard range [353.4, 406.6) as the reference voltage value, calculate the difference between the peak voltage and the reference voltage in the curve.

[0059] If the voltage rise is determined to be greater than 35V and the similarity coefficient is greater than 0.75, ensuring that the phenomenon occurs stably and is a typical photovoltaic scenario, then it is identified as the "0 < photovoltaic power generation < transformer area power consumption" scenario; otherwise, it is identified as the "0 = photovoltaic power generation < transformer area power consumption" scenario.

[0060] In one optional implementation, after identifying the photovoltaic scenario based on the gateway data, the process may further include: By combining weather data, the impact of overvoltage on users and equipment is statistically analyzed to further determine whether there are other factors that may lead to misjudgment of the information, while giving a confidence level.

[0061] Since the PMS system currently lacks weather information, this criterion can only be established by importing weather data. By selecting and determining the corresponding threshold information for rainy and sunny days based on the weather data of the station area, the confidence of the above criterion can be strengthened. If no weather data can be imported, this refined criterion can be skipped.

[0062] (1) Photovoltaic power generation > power consumption of the distribution area Enhanced Sunny Day Criterion: When the imported weather data indicates a sunny day, the light intensity during the preset time period is set. 500W / m 2 Photovoltaic output increases by 40%-60% compared to cloudy days. However, if the reverse current at the switching point continues... If the voltage rises by more than 35V within 15 minutes, the confidence level of the criterion increases to 90% (compared to 75% for conventional criterions). For example, in a certain distribution area, when the photovoltaic backfeed power reaches 50kW on a sunny day, the reverse current increases by 25A compared to a cloudy day, and the proportion of users experiencing overvoltage rises from 30% to 50%.

[0063] Misjudgment and Avoidance of Rainy Day Light Intensity: 200W / m 2In the event of reverse current at the distribution transformer, it is necessary to verify the data by combining it with the output data of the photovoltaic inverter (if available) or the data from a sunny day the following day, to avoid misjudgment of falsely high voltage caused by the transformer being set too high (e.g., level 4). In a certain mountainous area, the transformer was set abnormally on a rainy day, and the normal load current was mistakenly judged as reverse photovoltaic current. This was corrected after comparing the data from a sunny day.

[0064] Seasonal load rate adjustment: In summer, air conditioning load increases the load rate of the transformer area by 15%-20%. The photovoltaic backfeed power needs to exceed the power consumption of the transformer area by more than 20% before reverse current will occur (e.g., when the power consumption of the transformer area is 80kW, the photovoltaic output needs to be >96kW). In winter, the power factor of electric heating load is low (<0.7). The reactive power compensation effect of photovoltaic backfeed may mask the reverse current characteristics. It is necessary to combine the voltage harmonic distortion rate (THD>5%) for auxiliary judgment.

[0065] (2) 0 < Photovoltaic power generation < Electricity consumption in the distribution area Sunny day: The degree of agreement between voltage fluctuation frequency and light change cycle (cloud cover cycle - 10 minutes) When the output power is 80% and the fluctuation range is 1-3V, the confidence level of the criterion increases to 85% (normally 70%); on cloudy days: the fluctuation range of photovoltaic output decreases. (10% of rated power), if voltage fluctuation is detected at this time 5V, it is necessary to check whether the misjudgment is caused by the automatic adjustment of the transformer tap (accounting for 30% of misjudgment cases on cloudy days).

[0066] During the spring farming season: the irrigation pump cluster starts up (the load rate briefly rises to 50%), which may offset the photovoltaic voltage rise effect. This requires adjusting the voltage fluctuation frequency (0.1-2Hz) and load rate. 30% of the duration "2-hour" dual condition judgment.

[0067] Autumn light load period: Load rate is generally low The impact of a 10% increase in three-phase imbalance on voltage deviation is 1.5 times greater than in summer (e.g., the voltage deviation of phase A increases from 5V to 8V). It is necessary to combine weather data to eliminate the combined disturbance of "PV output fluctuation + imbalance".

[0068] S4. Based on the aforementioned data, perform transformer load rate analysis, transformer three-phase imbalance analysis, transformer power factor analysis, transformer tap position analysis, and transformer supply radius analysis to obtain the load rate analysis results, three-phase imbalance analysis results, power factor analysis results, tap position analysis results, and supply radius analysis results.

[0069] Among them, the transformer substation topology, as a structural factor affecting voltage quality, directly determines the voltage loss and stability during transmission and distribution through its line parameters, layout, and equipment configuration. The main key factors are analyzed as follows: Wire diameter: Wire diameter is a core parameter affecting voltage loss. According to the law of resistance, the resistance of a wire is inversely proportional to its diameter (for the same material and length); the smaller the diameter, the greater the resistance. In low-voltage power transmission, voltage loss is mainly manifested as the voltage drop corresponding to active power loss. U=IR, where I is the line current and R is the line resistance. When the wire diameter is insufficient, even under normal load, the line resistance will cause significant voltage drop, especially during peak load periods. Large currents flowing through thin conductors can cause the voltage at the end-user to drop significantly below the rated value, creating a "low voltage" problem. For example, some older transformer substations still use 16mm² wires. 2 For aluminum core wires of 200V and below, the voltage at the end may drop below 200V during peak summer air conditioning loads, seriously affecting the normal operation of electrical equipment.

[0070] Power supply radius: The power supply radius is significantly negatively correlated with voltage quality. When the power supply radius is too large, the total resistance of the line accumulates with the increase in length, and the voltage loss increases linearly. For low-voltage distribution areas, it is generally recommended that the reasonable power supply radius of a three-phase four-wire line be controlled within 500 meters. Beyond this range, voltage quality problems will deteriorate sharply. On the one hand, the voltage at the end users will remain low, making it difficult to meet the allowable deviation requirement of 220V±7% in the "Low-voltage Distribution Design Code"; on the other hand, long-distance lines are susceptible to environmental factors (such as temperature and humidity), resulting in decreased resistance stability. Furthermore, due to the increase in line distributed capacitance, a slight increase in end voltage may occur under light load, creating a potential "voltage fluctuation" hazard.

[0071] Yyn0 type distribution transformer: The impact of a Yyn0 connection distribution transformer (high-voltage side delta connection, low-voltage side star connection, and neutral point directly grounded) on voltage quality is reflected in its three-phase balance regulation capability. This connection method can effectively suppress zero-sequence current and stabilize phase voltage through neutral point grounding, but its compensation capability is limited when the three-phase load is unbalanced. When the load of one phase is much greater than that of the other two phases, the neutral point will shift, resulting in a decrease in the voltage of the heavily loaded phase and an increase in the voltage of the lightly loaded phase. For example, in a certain distribution area, when the load of phase A is 80kW and the loads of phases B and C are only 20kW, the voltage of phase A may drop to 190V, while the voltage of phase C may rise to 240V, both exceeding the allowable deviation range. This phenomenon is particularly common in rural distribution areas.

[0072] Low average capacity per household: Insufficient average transformer capacity per household is a significant contributing factor to voltage quality deterioration under load growth. Average capacity per household reflects the matching degree between the power supply capacity of the distribution area and the user load. When the average capacity per household is below 0.5kVA, the transformer is prone to overload during peak load periods. When a transformer is overloaded, its leakage impedance voltage drop (… U=I×(R k +jX k ), where Rk X k The increased short-circuit resistance and reactance lead to a decrease in the secondary output voltage. Simultaneously, overload causes the transformer temperature to rise, reducing insulation performance and further affecting its voltage regulation stability. With the increasing prevalence of high-power equipment (such as electric heating and charging stations for new energy vehicles) in residential electricity loads, the contradiction of insufficient capacity per household is becoming increasingly prominent. In some areas, during peak winter electricity consumption, the transformer load rate exceeds 120%, and the terminal voltage remains persistently low.

[0073] In addition, the three-phase load imbalance and line aging in the transformer substation topology can also indirectly affect voltage quality. When the three-phase imbalance exceeds 15%, it will exacerbate the phase voltage deviation; and aging lines, due to conductor oxidation and cross-sectional reduction, have increased resistance and voltage loss that is more than 30% higher than that of new lines, becoming a hidden contributor to potential voltage quality problems.

[0074] The load factor analysis based on the aforementioned data points yields the following results: If the maximum daily load rate of the distribution transformer exceeds the first preset percentage, the load rate analysis result is identified as a risk of distribution transformer overload. Obtain the number of households in the transformer substation area, and calculate the average capacity per household based on the number of households in the transformer substation area and the transformer capacity; If the average capacity per household is less than the preset capacity, and the average daily load rate of the distribution transformer exceeds the second preset percentage, then the load rate analysis result is identified as low average capacity per household exacerbating the voltage over-limit problem (overload leads to increased voltage loss).

[0075] In one optional implementation, the first preset percentage is 80%, the second preset percentage is 30%, and the preset capacity is 0.5 kVA.

[0076] In one optional implementation, after performing distribution transformer load rate analysis based on the gate data and obtaining the load rate analysis results, the process may further include: Under overload conditions, based on the relationship between line impedance and current ( (U=IR) Analyze the trend of voltage loss as the load rate increases (voltage loss increases by 5%-8% for every 10% increase in load rate), count the voltage over-limit situations of users during overload periods, and clarify the impact of load rate on voltage quality.

[0077] Among them, the three-phase imbalance analysis of the distribution transformer based on the aforementioned threshold data yields the following results: Obtain the transformer wiring type; If the transformer wiring type is Yyn0 wiring, and the maximum daily three-phase imbalance exceeds the third preset percentage, and the transformer daily average load rate exceeds the second preset percentage, then the three-phase imbalance analysis result is identified as a phase voltage deviation problem caused by three-phase imbalance. If the transformer wiring type is Dyn11 wiring, and the maximum daily three-phase imbalance exceeds the fourth preset percentage, and the transformer's daily average load rate exceeds the second preset percentage (when it exceeds the second preset percentage, the imbalance has a more significant impact on the voltage), then the three-phase imbalance analysis result is identified as a phase voltage deviation problem caused by the three-phase imbalance (the voltage of the heavier phase decreases, and the voltage of the lighter phase increases).

[0078] In one optional implementation, the third preset percentage is 15% and the fourth preset percentage is 25%.

[0079] In one alternative implementation, it may further include: To address the issue of inaccurate transformer registration information, a daily assessment is performed: Three-phase (A, B, C) voltage data at each transformer outlet are screened, and the maximum and minimum values ​​of the three-phase voltages at each moment are extracted. The phase voltage difference (maximum value - minimum value) is calculated. A transformer is identified as a suspected Yyn0 transformer if the following conditions are met simultaneously: (1) Phase voltage difference 20V; (2) Phase voltage difference 20V cumulative duration 5 hours.

[0080] By monitoring the three-phase voltage difference and its duration daily, the characteristics of the Yyn0 transformer can be accurately identified, providing a basis for subsequent three-phase imbalance management and voltage deviation optimization.

[0081] In one alternative implementation, it may further include: Statistically analyze the three-phase voltage deviation (maximum phase voltage - minimum phase voltage), analyze the linear relationship between unbalance and voltage deviation (for every 10% increase in unbalance, the phase voltage deviation increases by 3-5V), and clarify the range of impact on the voltage quality of the user side (e.g., unbalance in a certain phase causes the voltage of users in that phase to be generally low).

[0082] The power factor analysis of the distribution transformer based on the aforementioned data yields the following results: If the power factor of the distribution transformer is lower than the preset power factor, and the daily average load rate of the distribution transformer exceeds the second preset percentage, then the power factor analysis result is identified as a low power factor of the distribution transformer.

[0083] In one alternative implementation, the preset power factor is 0.9.

[0084] There may be capacitor compensation equipment in the transformer substation. When the user load is running, its voltage does not exceed the limit. However, when the user stops working, its capacitor equipment is not disconnected in time, causing its voltage to exceed the limit. This scenario is usually related to voltage fluctuations in the transformer substation. Therefore, the correlation between power factor fluctuations and voltage fluctuations needs to be considered.

[0085] In one alternative implementation, it may further include: When the power factor of the distribution transformer is lower than the preset value, the reactive current increases, leading to voltage loss. U=Q X / U (Q is reactive power) increases; statistical analysis of user voltage deviation in this scenario (e.g., the terminal voltage is 5-10V lower than the cut-off voltage) clarifies the effect of reactive power loss on voltage quality.

[0086] Among them, the gear position analysis based on the gate data yields the following results: Extract one week's worth of gate voltage data from the three-phase voltage values ​​at the gate; Based on the voltage data at the control points, the duration of severe overvoltage and undervoltage is superimposed to clarify the distribution of voltage over-limit issues before gear adjustment; After lowering the transformer speed, ensure that there are no new low-voltage users, and use the original number of overvoltage users as a base. If the number of overvoltage users decreases by more than the fifth preset percentage, the transformer speed adjustment is deemed effective. After adjusting the transformer speed, the number of users with severe overvoltage did not increase significantly. Based on the original number of low-voltage users, if the number of low-voltage users needs to be reduced by more than the sixth preset percentage, then the transformer speed adjustment is deemed reasonable.

[0087] In one optional implementation, the percentage of valid monitoring days for the continuous week's threshold voltage data, after excluding invalid data days such as equipment failures and power outages, should reach 80%-90%. Severe overvoltage is defined as a voltage greater than 253V, undervoltage as a voltage less than 198V, and overvoltage as a voltage greater than 235.4V. The fifth preset percentage is 20%. The sixth preset percentage is 20%.

[0088] The power supply radius analysis based on the gateway data yields the following results: Calculate the power supply radius of the transformer substation; Based on the aforementioned transformer area topology and user distribution GIS data, the power supply area type and load density characteristics are determined. The power supply radius threshold is matched according to the power supply area type and the load density characteristics, as shown in Table 2; The power supply radius of the transformer area is compared with the power supply radius threshold to obtain the power supply radius analysis results.

[0089] Specifically, if the power supply radius of the transformer area is greater than the power supply radius threshold, the power supply radius analysis result is determined to be that the power supply radius is too long.

[0090] The power supply radius of the transformer substation is the electrical distance from the low-voltage side outgoing terminal of the transformer to the electricity meter of the farthest user in the substation (not the geographical straight-line distance, and the route and laying method of the line need to be taken into account).

[0091] In one optional implementation, the power supply radius of the transformer area is calculated by accumulating the line length and correcting it with a laying coefficient. For example, the cable laying coefficient is taken as 1.05-1.1, and the overhead line coefficient is taken as 1.1-1.2 to compensate for geographical deviation.

[0092] Table 2 Power Supply Radius Threshold Matching Rules

[0093] S5. Using a decision tree model, perform multi-factor correlation analysis on the photovoltaic scenario, the load rate analysis results, the three-phase imbalance analysis results, the power factor analysis results, the tap analysis results, and the power supply radius analysis results to obtain the contribution ratio of each factor to voltage exceedance, specifically including S51-S54: S51. Prioritize the impact of the photovoltaic scenario, the load rate analysis result, the three-phase imbalance analysis result, the power factor analysis result, the gear analysis result, and the power supply radius analysis result.

[0094] Specifically, based on "coverage of users + severity of voltage exceedance", the priorities from highest to lowest are: substation-level issues (such as excessively long power supply radius, unreasonable transformer taps). Three-phase system problems (such as three-phase imbalance, transformer overload) Localized problems (such as fluctuations in photovoltaic output at a single point or low local power factor).

[0095] S52. Determine the criterion confidence levels of the photovoltaic scenario, the load rate analysis result, the three-phase imbalance analysis result, the power factor analysis result, the gear analysis result, and the power supply radius analysis result.

[0096] In one alternative implementation, it may further include: Adjust the confidence level of the criterion.

[0097] Specifically, among reasons of the same level, they are sorted in a second order according to the confidence level of the criterion (e.g., the confidence level of the criterion is high for the sunny solar photovoltaic scenario). Conventional criterion confidence level, transmission gear shifting verification passed item Unverified items) were removed, along with unreliable results with low confidence levels (such as detailed photovoltaic criteria without weather data support and power factor analysis with missing data).

[0098] S53. Sort the photovoltaic scenario, the load rate analysis result, the three-phase imbalance analysis result, the power factor analysis result, the gear analysis result, and the power supply radius analysis result according to the priority of the degree of influence and the confidence level of the criterion to obtain a list of possible causes of the transformer area voltage exceeding the limit problem.

[0099] S54. Input the list of possible causes of the voltage over-limit problem in the transformer area into the decision tree model for multi-factor correlation analysis, and output the contribution ratio of each factor to the voltage over-limit problem.

[0100] The decision tree model performs the following operations: Data correlation verification matches the coupling relationship between results from different dimensions: For example, if both "photovoltaic backfeed" and "transformer tap too high" exist in the list, the superposition effect of the two needs to be analyzed to avoid independent attribution; if both "three-phase imbalance" and "power supply radius too long" exist, the synergistic effect of the two on voltage deviation needs to be derived by combining relevant principles.

[0101] Conflict resolution is achieved when different criteria contradict each other, with decisions based on "data integrity + scenario adaptability": for example, in a photovoltaic scenario, the criteria are determined by "photovoltaic power generation capacity". The power consumption figure shows that the transformer load rate analysis indicates overload. The decision tree model needs to combine weather data to statistically analyze the impact of overvoltage on users and equipment, and further determine whether there are other factors that may cause misjudgment of the information. At the same time, a confidence level step is given to determine which conclusion to accept and to mark possible interference factors.

[0102] Impact weight allocation: Based on the aforementioned analysis, a weight model is established to comprehensively consider the degree of influence of each factor on the voltage limit problem, and the factors are sorted according to the total weight to clarify the contribution ratio of each cause to the voltage limit problem.

[0103] This invention utilizes decision trees for multi-factor fusion, with its core strength lying in its hierarchical branching structure, which perfectly aligns with the complex "multiple causes, one effect" problem of voltage exceeding limits in transformer substations. This model can systematically decompose the coupling relationships between multiple factors such as photovoltaics, topology, and load, and, thanks to its white-box characteristics, provides clear and interpretable attribution paths, greatly enhancing the credibility and reliability of the criteria. Simultaneously, decision trees effectively utilize limited indirect measurement data such as voltage and current to deduce key information like power relationships through rule-based reasoning, solving the challenge of judgment under data shortages. Ultimately, the model outputs structured and prioritized causal conclusions, directly supporting the formulation of differentiated and actionable operation and maintenance strategies. This overcomes the problems of simple criteria and high misjudgment rates in traditional decision tree algorithms (expert systems), making it a core technical means for achieving accurate identification and attribution.

[0104] In one alternative implementation, it further includes: S6. Based on the contribution ratio of each factor to voltage over-limit, a voltage over-limit diagnostic report is generated using knowledge graphs and large language models.

[0105] Specifically, based on the contribution ratio of each factor to voltage exceedance, a knowledge graph is used for correlation verification, and a weighted model and a large language model are used for influence allocation and ranking. Finally, the large language model automatically generates a voltage exceedance diagnostic report that includes the core voltage problem, the reasons for the multi-factor fusion, priority governance suggestions, and an explanation of data limitations. As shown in Table 3, Table 3 shows the report specifications that can be referenced.

[0106] Table 3 Report Specifications

[0107] like Figure 4 As shown, according to Figure 4 The voltage over-limit diagnostic report generated from the test bench topology diagram shown is as follows: District Number: Township 10kV- Line 01 transformer area (transformer capacity 400kVA, 86 users, Class D area); Core voltage issue: Continuous overvoltage during specific periods, affecting a certain percentage of users; Reasons for the fusion of multiple factors: 1. Inappropriate transformer speed setting: Data verification showed that the original speed setting caused overvoltage. After lowering the speed setting, the number of users experiencing overvoltage decreased, confirming that the excessively high speed setting was the direct cause. 2. Photovoltaic backfeed scenario: Photovoltaic backfeed occurs during specific periods on sunny days. The criteria have a high confidence level. The combined effect of photovoltaic backfeed and the range of photovoltaic outputs exacerbates the overvoltage. 3. Three-phase imbalance: The distribution transformer has a three-phase imbalance problem, with the voltage of the lighter phase being higher than that of the heavier phase, which further expands the overvoltage range; Priority management recommendations: 1. Prioritize adjusting the transformer tap level; 2. Verify the three-phase load distribution and adjust the single-phase user access phase; 3. Assess the photovoltaic access capacity and consider installing reactive power compensation devices.

[0108] Decision-makers can base their decisions on Figure 4 The topology diagram and the method described above in this invention are used to reconfirm and judge the voltage over-limit diagnosis report.

[0109] The present invention provides a low-voltage distribution area voltage over-limit identification method. Based on collected user voltage data, it identifies whether a voltage over-limit exists and obtains the identification result. If a voltage over-limit is identified, it identifies the photovoltaic scenario based on collected gate data. Based on the gate data, it performs distribution transformer load rate analysis, distribution transformer three-phase imbalance analysis, distribution transformer power factor analysis, distribution transformer tap analysis, and distribution area power supply radius analysis, obtaining load rate analysis results, three-phase imbalance analysis results, power factor analysis results, tap analysis results, and power supply radius analysis results. A decision tree model is then used to analyze the photovoltaic scenario, load rate analysis results, and three-phase imbalance analysis results. Multi-factor correlation analysis was conducted using the analysis results of power factor analysis, voltage level analysis, and power supply radius analysis to obtain the contribution ratio of each factor to voltage exceedance. Using this data and a decision tree model, a comprehensive judgment and accurate attribution of multiple factors were achieved when voltage exceedances were identified. Through the hierarchical classification logic of the decision tree, the contribution of each factor to voltage exceedances was quantitatively analyzed, clearly identifying the dominant cause of voltage exceedances and avoiding misattribution to photovoltaics. This improved the accuracy and comprehensiveness of voltage exceedance cause identification, providing solid support for developing differentiated and effective governance strategies. Furthermore, based on accurate cause analysis, differentiated solutions can be provided for different types of voltage exceedance problems. For example, if the problem is determined to be caused by excess photovoltaic power, photovoltaic output regulation can be optimized or local consumption can be guided; if it is caused by three-phase imbalance, load distribution can be adjusted or balancing devices can be installed. This reduces the waste of manpower, material resources, and financial resources caused by misjudgment, improves the effectiveness and economy of governance measures, and promotes the coordinated and efficient operation of distributed photovoltaics and the distribution network.

[0110] According to another aspect of the invention, Figure 2 This is a schematic diagram illustrating a low-voltage zone voltage over-limit identification device according to an embodiment of the present invention. The device includes: The data acquisition module is used to collect user voltage data and gate data in the low-voltage distribution area; The voltage limit exceedance detection module is used to identify whether there is a voltage limit exceedance based on the user voltage data and obtain the detection result; A photovoltaic scene identification module is used to identify a photovoltaic scene based on the threshold data if the identification result indicates that there is a voltage over-limit. The factor analysis module is used to perform distribution transformer load rate analysis, distribution transformer three-phase imbalance analysis, distribution transformer power factor analysis, distribution transformer tap analysis, and distribution area power supply radius analysis based on the aforementioned data, and to obtain load rate analysis results, three-phase imbalance analysis results, power factor analysis results, tap analysis results, and power supply radius analysis results. The multi-factor correlation analysis module is used to perform multi-factor correlation analysis on the photovoltaic scenario, the load rate analysis results, the three-phase imbalance analysis results, the power factor analysis results, the tap analysis results, and the power supply radius analysis results using a decision tree model, so as to obtain the contribution ratio of each factor to voltage over-limit.

[0111] In one optional implementation, the threshold data includes the threshold three-phase voltage value and the threshold three-phase current value; identifying the photovoltaic scenario based on the threshold data includes: determining whether the threshold three-phase current value is reversed; if it is reversed, the photovoltaic scenario is identified as a scenario where the photovoltaic power generation is greater than the power consumption of the transformer substation; if it is not reversed, the historical data of the threshold three-phase voltage value during the same period of a preset time period is compared to determine whether the threshold three-phase voltage value has abnormally increased; if it is, the photovoltaic scenario is identified as a scenario where the power consumption of the transformer substation is greater than the photovoltaic power generation (which is greater than zero); if it is not, the photovoltaic scenario is identified as a scenario where the power consumption of the transformer substation is greater than the photovoltaic power generation (which is equal to zero).

[0112] In one optional implementation, the threshold data includes the daily maximum load rate of the distribution transformer, the daily average load rate of the distribution transformer, and the distribution transformer capacity. Based on the threshold data, the distribution transformer load rate analysis is performed to obtain the load rate analysis results, including: if the daily maximum load rate of the distribution transformer exceeds a first preset percentage, the load rate analysis result is identified as a risk of distribution transformer overload; the number of households in the transformer area is obtained, and the average capacity per household is calculated based on the number of households in the transformer area and the distribution transformer capacity; if the average capacity per household is less than a preset capacity, and the daily average load rate of the distribution transformer exceeds a second preset percentage, the load rate analysis result is identified as a low average capacity per household exacerbating the voltage over-limit problem.

[0113] In one optional implementation, the threshold data further includes the daily maximum three-phase imbalance; based on the threshold data, a three-phase imbalance analysis of the distribution transformer is performed to obtain the three-phase imbalance analysis result, including: obtaining the distribution transformer connection type; if the distribution transformer connection type is Yyn0 connection, and the daily maximum three-phase imbalance exceeds a third preset percentage, and the daily average load rate of the distribution transformer exceeds a second preset percentage, then the three-phase imbalance analysis result is identified as a phase voltage deviation problem caused by three-phase imbalance; if the distribution transformer connection type is Dyn11 connection, and the daily maximum three-phase imbalance exceeds a fourth preset percentage, and the daily average load rate of the distribution transformer exceeds a second preset percentage, then the three-phase imbalance analysis result is identified as a phase voltage deviation problem caused by three-phase imbalance.

[0114] In one optional implementation, the threshold data further includes the distribution transformer power factor; based on the threshold data, the distribution transformer power factor analysis is performed to obtain the power factor analysis result, including: if the distribution transformer power factor is lower than a preset power factor and the daily average load rate of the distribution transformer exceeds a second preset percentage, then the power factor analysis result is identified as the distribution transformer power factor being too low.

[0115] In one optional implementation, transformer tap position analysis is performed based on the threshold data to obtain the tap position analysis results, including: extracting one week's worth of threshold voltage data from the three-phase voltage values ​​at the threshold; determining the distribution of voltage exceedance issues before tap position adjustment by superimposing the threshold voltage data with the duration of severe overvoltage and undervoltage; after lowering the transformer tap position, ensuring no new undervoltage users are added, and using the original number of overvoltage users as a base, if the number of overvoltage users decreases by more than a fifth preset percentage, then the transformer tap position adjustment is deemed effective; after raising the transformer tap position, if the number of severe overvoltage users does not increase significantly, and using the original number of undervoltage users as a base, if the number of undervoltage users needs to be reduced by more than a sixth preset percentage, then the transformer tap position adjustment is deemed reasonable.

[0116] In one optional implementation, the gateway data further includes transformer area topology and user distribution GIS data; the power supply radius analysis of the transformer area is performed based on the gateway data to obtain the power supply radius analysis result, including: calculating the power supply radius of the transformer area; determining the power supply area type and load density characteristics based on the transformer area topology and user distribution GIS data; matching the power supply radius threshold according to the power supply area type and the load density characteristics; and comparing the transformer area power supply radius with the power supply radius threshold to obtain the power supply radius analysis result.

[0117] In one optional implementation, a decision tree model is used to perform multi-factor correlation analysis on the photovoltaic scenario, the load factor analysis results, the three-phase imbalance analysis results, the power factor analysis results, the tap position analysis results, and the power supply radius analysis results to obtain the contribution ratio of each factor to voltage exceedance. This includes: prioritizing the influence of the photovoltaic scenario, the load factor analysis results, the three-phase imbalance analysis results, the power factor analysis results, the tap position analysis results, and the power supply radius analysis results; determining the criterion confidence level of the photovoltaic scenario, the load factor analysis results, the three-phase imbalance analysis results, the power factor analysis results, the tap position analysis results, and the power supply radius analysis results; sorting the photovoltaic scenario, the load factor analysis results, the three-phase imbalance analysis results, the power factor analysis results, the tap position analysis results, and the power supply radius analysis results according to the influence priority and the criterion confidence level to obtain a list of possible causes of voltage exceedance in the transformer area; and inputting the list of possible causes of voltage exceedance in the transformer area into the decision tree model for multi-factor correlation analysis to output the contribution ratio of each factor to voltage exceedance.

[0118] In one alternative implementation, the system further includes a report generation module for generating a voltage over-limit diagnostic report using a knowledge graph and a large language model based on the contribution percentage of each factor to the voltage over-limit.

[0119] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of this application can be implemented in various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0120] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0121] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0122] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0123] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0124] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A low-voltage transformer area voltage out-of-limit identification method, characterized in that, The method comprises the steps of: collecting user voltage data and gateway data of a low-voltage transformer area; identifying whether there is voltage overrun according to the user voltage data to obtain an identification result; if the identification result is that there is voltage overrun, identifying a photovoltaic scene based on the gateway data; performing distribution transformer load rate analysis, distribution transformer three-phase imbalance analysis, distribution transformer power factor analysis, distribution transformer grade analysis, and transformer area power supply radius analysis based on the gateway data to obtain load rate analysis results, three-phase imbalance analysis results, power factor analysis results, grade analysis results, and power supply radius analysis results; performing multi-factor correlation analysis on the photovoltaic scene, the load rate analysis results, the three-phase imbalance analysis results, the power factor analysis results, the grade analysis results, and the power supply radius analysis results using a decision tree model to obtain a contribution ratio of each factor to voltage overrun.

2. The low-voltage area voltage out-of-limit identification method according to claim 1, characterized in that, The gateway data comprises gateway three-phase voltage values and gateway three-phase current values; identifying the photovoltaic scene based on the gateway data comprises: determining whether the gateway three-phase current values are reversed, if reversed, identifying the photovoltaic scene as a scene in which photovoltaic power generation power is greater than transformer area power consumption power, if not reversed, comparing historical simultaneous period data of the gateway three-phase voltage values of a preset period to determine whether the gateway three-phase voltage values are abnormally lifted, if yes, identifying the photovoltaic scene as a scene in which transformer area power consumption power is greater than photovoltaic power generation power greater than zero, if no, identifying the photovoltaic scene as a scene in which transformer area power consumption power is greater than photovoltaic power generation power equal to zero.

3. The low-voltage area voltage overrun identification method of claim 1, wherein, The gateway data comprises distribution transformer daily maximum load rate, distribution transformer daily average load rate, and distribution transformer capacity; performing distribution transformer load rate analysis based on the gateway data to obtain load rate analysis results comprises: if the distribution transformer daily maximum load rate exceeds a first preset percentage, identifying the load rate analysis result as a distribution transformer overload risk; obtaining the number of households in the transformer area, and calculating household capacity according to the number of households in the transformer area and the distribution transformer capacity; if the household capacity is less than a preset capacity and the distribution transformer daily average load rate exceeds a second preset percentage, identifying the load rate analysis result as a problem of low household capacity aggravating voltage overrun.

4. The low-voltage area voltage overrun identification method according to claim 3, characterized in that, The gateway data further comprises daily maximum three-phase imbalance degree; performing distribution transformer three-phase imbalance analysis based on the gateway data to obtain three-phase imbalance analysis results comprises: obtaining the distribution transformer connection type; if the distribution transformer connection type is Yyn0 connection and the daily maximum three-phase imbalance degree exceeds a third preset percentage, and the distribution transformer daily average load rate exceeds a second preset percentage, identifying the three-phase imbalance analysis result as a phase voltage deviation problem caused by three-phase imbalance; if the distribution transformer connection type is Dyn11 connection and the daily maximum three-phase imbalance degree exceeds a fourth preset percentage, and the distribution transformer daily average load rate exceeds a second preset percentage, identifying the three-phase imbalance analysis result as a phase voltage deviation problem caused by three-phase imbalance.

5. The low-voltage area voltage overrun identification method according to claim 3, characterized in that, The gateway data further comprises distribution transformer power factor; performing distribution transformer power factor analysis based on the gateway data to obtain power factor analysis results comprises: If the distribution transformer power factor is lower than the preset power factor, and the daily average load rate of the distribution transformer exceeds the second preset percentage, the power factor analysis result is identified as low distribution transformer power factor.

6. The low-voltage area voltage out-of-limit identification method according to claim 2, characterized in that, Based on the key data, distribution transformer tap position analysis is performed to obtain tap position analysis results, including: Extracting key voltage data for a continuous week from the key three-phase voltage values; Based on the key voltage data, superimposing the duration of serious overvoltage and low voltage, the voltage out-of-limit problem distribution before tap position adjustment is determined; After adjusting the distribution transformer tap position downward, it is ensured that there is no new low voltage user, and based on the original number of overvoltage users, if the number of overvoltage users is reduced by more than the fifth preset percentage, it is determined that the distribution transformer tap position adjustment is effective; After adjusting the distribution transformer tap position upward, there is no significant increase in the number of serious overvoltage users, and based on the original number of low voltage users, if the number of low voltage users is reduced by more than the sixth preset percentage, it is determined that the distribution transformer tap position adjustment is reasonable.

7. The low-voltage transformer area voltage overrun identification method of claim 1, wherein, The key data also includes GIS data of the distribution area topology and user distribution; Based on the key data, distribution area power supply radius analysis is performed to obtain power supply radius analysis results, including: Calculating the distribution area power supply radius; Based on the distribution area topology and user distribution GIS data, determine the power supply area type and load density characteristics; According to the power supply area type and the load density characteristics, match the power supply radius threshold value; Compare the distribution area power supply radius with the power supply radius threshold value to obtain the power supply radius analysis result.

8. The low-voltage transformer area voltage overrun identification method of claim 1, wherein, Using a decision tree model to perform multi-factor correlation analysis on the photovoltaic scene, the load rate analysis result, the three-phase imbalance analysis result, the power factor analysis result, the tap position analysis result, and the power supply radius analysis result to obtain the contribution ratio of each factor to voltage out-of-limit, including: Divide the influence degree priority of the photovoltaic scene, the load rate analysis result, the three-phase imbalance analysis result, the power factor analysis result, the tap position analysis result, and the power supply radius analysis result; Determine the criterion confidence of the photovoltaic scene, the load rate analysis result, the three-phase imbalance analysis result, the power factor analysis result, the tap position analysis result, and the power supply radius analysis result; According to the influence degree priority and the criterion confidence, sort the photovoltaic scene, the load rate analysis result, the three-phase imbalance analysis result, the power factor analysis result, the tap position analysis result, and the power supply radius analysis result to obtain a list of possible causes of distribution area voltage out-of-limit problems; Input the list of possible causes of distribution area voltage out-of-limit problems into a decision tree model for multi-factor correlation analysis, and output the contribution ratio of each factor to voltage out-of-limit.

9. The low-voltage transformer area voltage overrun identification method of claim 1, wherein, Also includes: Based on the contribution ratio of each factor to voltage out-of-limit, use a knowledge graph and a large language model to generate a voltage out-of-limit diagnosis report.

10. A low-voltage transformer area voltage out-of-limit identification device, characterized in that, It includes: A data acquisition module for acquiring user voltage data and key data of a low-voltage distribution area; A voltage out-of-limit identification module for identifying whether there is voltage out-of-limit based on the user voltage data to obtain an identification result; A photovoltaic scene identification module for identifying a photovoltaic scene based on the key data if the identification result is voltage out-of-limit. a factor analysis module, configured to perform distribution transformer load rate analysis, distribution transformer three-phase imbalance analysis, distribution transformer power factor analysis, distribution transformer grade analysis, and distribution area power supply radius analysis based on the gateway data, to obtain load rate analysis results, three-phase imbalance analysis results, power factor analysis results, grade analysis results, and power supply radius analysis results; a multi-factor correlation analysis module, configured to perform multi-factor correlation analysis on the photovoltaic scene, the load rate analysis results, the three-phase imbalance analysis results, the power factor analysis results, the grade analysis results, and the power supply radius analysis results using a decision tree model, to obtain a contribution proportion of each factor to voltage out-of-limit.

Images