Distributed photovoltaic uploading risk early warning method and device and computer readable storage medium

Abstract

The invention discloses a distributed photovoltaic uploading risk early warning method and device and a computer readable storage medium, and the method comprises the steps: collecting power grid operation data and distributed photovoltaic data, and determining a distributed photovoltaic output prediction value and a load demand prediction value according to the power grid operation data and the distributed photovoltaic data; according to the distributed photovoltaic output prediction value and the load demand prediction value, determining a voltage safety risk, a power flow safety risk and a frequency safety risk after the power grid is accessed to the distributed photovoltaic; matching a corresponding risk application scene based on the voltage safety risk, the power flow safety risk and the frequency safety risk, and determining a corresponding early warning level according to the risk application scene; the distributed photovoltaic uploading risk is comprehensively considered from multiple dimensions of voltage safety, power flow safety and frequency safety, quantification of the distributed photovoltaic uploading risk is achieved, the accuracy and reliability of risk early warning are improved, and the fine requirement of power grid dispatching can be met.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a method, device, and computer-readable storage medium for early warning of risks in distributed photovoltaic power transmission. Background Technology

[0002] To address climate change, building a new power system has become a core task for my country to achieve energy transition and sustainable development. The new power system prioritizes energy and power security, with the construction of a high-proportion renewable energy supply and consumption system as its main task. It aims to achieve coordinated development of efficient energy utilization and low-carbon transformation by improving the system's flexibility, reliability, and intelligence. Distributed photovoltaic power generation, with its significant advantages such as being green, clean, and flexible in deployment, has ushered in a golden age of rapid development amidst the transformation wave, with its capacity continuously achieving new breakthroughs.

[0003] However, with the large number of distributed photovoltaic projects connected to the grid, their proportion in the energy structure continues to increase, and the scope and depth of their impact on the operation of the power system continue to increase. (1) The difficulty of grid operation and control increases. Distributed photovoltaics have the characteristics of randomness, volatility and intermittency. After large-scale connection, the grid has difficulty in accurately grasping the changes in grid-connected power and the growth of power load, and the difficulty of dispatching work increases. (2) Power quality is affected. After the connection of distributed photovoltaics, the voltage distribution of the grid will change, resulting in voltage rise in some areas, and even voltage exceeding the limit. (3) The grid's frequency and voltage regulation capabilities decrease. After the large-scale grid connection of distributed photovoltaics, the start-up space of conventional units is squeezed, the system's rotational inertia and damping decrease, and the frequency and voltage regulation capabilities continue to decline. During faults, distributed photovoltaics cannot provide effective reactive power support, and the transient voltage level decreases.

[0004] Currently, domestic research on large-scale grid integration of distributed photovoltaic (PV) power has yielded certain results. First, research focuses on grid carrying capacity and absorption capacity. Addressing the challenges posed by large-scale PV integration to grid carrying capacity and absorption capacity, research is conducted on the grid integration and absorption capacity of PV under different penetration rates. Mathematical models are established to analyze the grid's capacity to accommodate PV. Second, PV power prediction is being implemented. Meteorological data, historical power data, and machine learning algorithms (such as the improved XGBoost algorithm and the OMO-IBIGRU model) are used to predict the short-term output power of PV, improving the grid's dispatch flexibility and reliability. Third, risk assessment and prevention strategies are being researched. Assessing the grid operation risks caused by PV is crucial for ensuring the safe and stable operation of the grid. Researchers employ improved Monte Carlo mixed sampling and semi-invariant methods to analyze the risk levels of the grid under different operating conditions and propose corresponding prevention strategies. Fourth, the rationality of energy storage system configuration is being studied. The application of energy storage systems in PV grid integration is being investigated. Through the charging and discharging control of energy storage systems, power fluctuations of PV can be mitigated, improving the grid's capacity to absorb PV.

[0005] Research on distributed photovoltaic (PV) grid connection mainly focuses on improving grid carrying capacity, power prediction accuracy, and risk assessment and control strategies. However, there is still room for improvement in risk early warning. Firstly, there is a lack of systematic, comprehensive, and quantifiable risk assessment systems. Existing research on distributed PV grid connection risk assessment largely emphasizes qualitative analysis, lacking clear quantification and classification of risks, which fails to meet the refined requirements of grid dispatch and operation. Secondly, there are shortcomings in research on multi-factor coupled risk early warning methods. The risks of distributed PV grid connection are influenced by multiple factors such as meteorological conditions, grid topology, and energy storage system configuration. However, existing research often focuses on the impact of a single factor on risk, lacking early warning methods for multi-factor coupled risks. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to provide a method, device and computer-readable storage medium for risk early warning of distributed photovoltaic power transmission, which can improve the accuracy and reliability of risk early warning.

[0007] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: A method for early warning of risks in distributed photovoltaic power transmission includes: Collect power grid operation data and distributed photovoltaic data, and determine the predicted output value and load demand value of distributed photovoltaic based on the power grid operation data and distributed photovoltaic data; Based on the predicted output and load demand of the distributed photovoltaic power generation, determine the voltage security risk, power flow security risk, and frequency security risk after the grid is connected to distributed photovoltaic power generation. Based on the voltage safety risk, power flow safety risk, and frequency safety risk, the corresponding risk application scenarios are matched, and the corresponding early warning level is determined according to the risk application scenarios.

[0008] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is as follows: A distributed photovoltaic power transmission risk warning device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the aforementioned distributed photovoltaic power transmission risk warning method.

[0009] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is as follows: A computer-readable storage medium storing computer program instructions thereon, which, when executed by a processor, implement the steps of the above-described distributed photovoltaic power transmission risk warning method.

[0010] The beneficial effects of this invention are as follows: First, based on grid operation data and distributed photovoltaic (PV) data, the predicted output and load demand of distributed PV are determined. Then, based on the predicted output and load demand, the voltage security risk, power flow security risk, and frequency security risk after the grid is connected to distributed PV are determined. Finally, based on the voltage security risk, power flow security risk, and frequency security risk, corresponding risk application scenarios are matched, and the corresponding early warning level is determined according to the risk application scenarios. By comprehensively considering the distributed PV transmission risk from multiple dimensions such as voltage security, power flow security, and frequency security, the risk of distributed PV transmission is quantified. After quantification, the corresponding risk application scenarios are further matched, and the corresponding early warning level is determined according to the risk application scenarios. By linking the quantification results with the actual application scenarios, the accuracy and reliability of risk early warning are improved, which can meet the refined requirements of grid dispatching. Attached Figure Description

[0011] Figure 1 This is a flowchart illustrating the steps of a distributed photovoltaic power transmission risk early warning method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the distributed photovoltaic grid connection structure according to an embodiment of the present invention; Figure 3 This is a schematic diagram of a distributed photovoltaic power transmission risk early warning device according to an embodiment of the present invention. Detailed Implementation

[0012] 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.

[0013] The distributed photovoltaic (PV) power transmission risk early warning method, device, and computer-readable storage medium described in this application are applicable to scenarios where distributed PV is used in power systems to improve the accuracy of early warning. The following detailed embodiments illustrate these methods: In one alternative implementation, such as Figure 1 As shown, a method for early warning of risks in distributed photovoltaic power transmission includes the following steps: S1. Collect power grid operation data and distributed photovoltaic data, and determine the predicted value of distributed photovoltaic power output and the predicted value of load demand based on the power grid operation data and distributed photovoltaic data. Among them, the power grid operation data includes data such as voltage, current, and power of each node in the power grid, as well as line parameters such as resistance, reactance, and rated capacity; The distributed photovoltaic data includes real-time power generation, inverter status, irradiance, temperature, and other data of the distributed photovoltaic power station; S2. Determine the voltage security risk, power flow security risk, and frequency security risk of grid connection after the distributed photovoltaic power generation is connected to the grid based on the distributed photovoltaic power output forecast and load demand forecast. S3. Based on the voltage safety risk, power flow safety risk and frequency safety risk, match the corresponding risk application scenarios, and determine the corresponding early warning level according to the risk application scenarios; Among them, the primary mapping relationship between risk application scenarios and early warning levels can be determined in advance; The corresponding early warning level is determined based on the risk application scenario and the first mapping relationship; In this embodiment, the distributed photovoltaic refers to distributed photovoltaic for natural persons’ households and distributed photovoltaic for non-natural persons’ households. The former is distributed photovoltaic with a voltage level of no more than 380V at the connection point with the public power grid and no more than 10 kV (20 kV) at the connection point with the public power grid, and the total installed capacity is no more than 6 MW. Both are connected to the corresponding nodes according to the location of the distributed photovoltaic development. The output range of distributed photovoltaic nodes and the output coefficient C ou Regarding this, based on the current operation of distributed photovoltaic (PV) projects, the maximum output of distributed PV systems generally reaches 80%-90% of the total installed capacity, but does not exceed the maximum installed capacity. If no output is generated at night, then 0 < C. ou <100%; This implementation method defines three wiring topologies for distributed photovoltaic (PV) systems: if all PV power is self-consumed, an anti-backflow device ensures no power is fed back to the grid; if all PV power is fed to the grid, a dedicated grid connection point is set up at the 10kV busbar of the public distribution network, and all electricity is directly fed into the grid via a step-up transformer / combiner box; if self-consumed power with surplus electricity is fed to the grid, the PV grid connection point is on the user side, and after bidirectional metering, surplus electricity is fed back to the distribution network via the same low-voltage busbar. The voltage levels of 10kV and below defined in this embodiment are calculated based on the principle that there is no reverse overload (≤80%) and the maximum new installed capacity can be consumed locally or gradually. This is also one of the important boundary conditions for early warning in this embodiment. Acceptable upper limit = distribution transformer capacity × 80% - PLmax.

[0014] In another alternative implementation, meteorological data collection is also included; The meteorological data includes parameters such as wind speed, rainfall, and cloud cover that indirectly affect the power generation of distributed photovoltaics. The meteorological data is reflected in the subsequent distributed photovoltaic power output prediction, and its impact on the total output can be considered when selecting prediction method f. The process of determining the predicted output and load demand of distributed photovoltaic power generation and the predicted load demand based on the power grid operation data and distributed photovoltaic data includes: The load demand forecast value is determined based on the power grid operation data; The predicted output value of distributed photovoltaic power is determined based on the distributed photovoltaic data and meteorological data. In practical implementation, for distributed photovoltaic power output prediction, the operating power output data of distributed photovoltaic power stations in a certain area is set as P. PV The installed capacity of distributed photovoltaic power is P inst The output coefficient is C ou The per-unit value is:

[0015] In the formula, P pu (i, d, t) represents the per-unit value at time t on day d of year i, P m (i, d, t) represents the actual output of distributed photovoltaic power at time t on day d of year i, P max (i, d) represents the maximum output of distributed photovoltaic power on day d of year i. The output data is historical data, which can be collected (generally obtained through system export or statistics from other sources).

[0016] In the formula, P pre(d, t) represents the predicted per-unit value at time t on day d in the future. The f function is a distributed photovoltaic power output prediction model, which can be predicted using existing statistical analysis methods, such as time series prediction. The per-unit values ​​are sorted by time to capture their growth trend (growth rate) and periodicity (spring, summer, autumn, winter, holidays, rain / sunny, etc.) for prediction. The per-unit values ​​at time t on the same day in the past ten years can be used as input to output the predicted per-unit values.

[0017]

[0018] In the formula, P PV (d, t) represents the predicted output of distributed photovoltaic power at time t on day d in the future.

[0019] For load demand forecasting, let the load demand in a certain area be P. L The system embeds a statistical model for prediction:

[0020] In the formula, P L (d, t) represents the load demand forecast result at time t on day d in the future, P L (i, d, t) represents the load demand at time t on day d of year i in history. This data is historical data and can be collected (usually obtained through system export or statistics from other sources). The f function is the load demand forecasting function, which can be predicted using existing statistical analysis methods.

[0021] In this embodiment, unlike traditional studies that isolate and analyze single factors, the influence of meteorological conditions on distributed photovoltaic power output is fully considered, ensuring the reliability and accuracy of the predicted distributed photovoltaic power output value.

[0022] In another alternative implementation, after collecting grid operation data and distributed photovoltaic data, the method further includes: For the collected data, the 3σ principle is used to identify whether it exceeds the preset range. If so, the data exceeding the preset range is removed, and the gaps are filled by linear interpolation. In practical implementation, if the collected data contains values ​​outside a reasonable range (0 or far exceeding the rated value), the 3σ principle is used for identification and correction. Assume the collected data sequence is {x1, x2, ..., x...} n The mean is μ, and the standard deviation is σ.

[0023]

[0024] The valid data range is set to [μ-3σ, μ+3σ]. After removing outliers, linear interpolation is used to fill in the gaps.

[0025] In this embodiment, by preprocessing the data, removing abnormal data and filling it with linear interpolation, the reliability of the data for subsequent statistical analysis and prediction is ensured, further guaranteeing the accuracy of the prediction.

[0026] In another alternative implementation, it further includes: The impact indicators corresponding to the voltage safety risk, power flow safety risk, and frequency safety risk are determined in advance. In this embodiment, the impact indicators corresponding to voltage safety risks may include: Node voltage deviation rate is used to measure the degree of deviation between the predicted voltage and the rated voltage after distributed photovoltaic grid connection. Node voltage fluctuation rate is used to measure the short-term fluctuation of line voltage after distributed photovoltaic (PV) grid connection. The magnitude of the change; The node voltage over-limit rate reflects the proportion of grid voltage exceeding the allowable range after distributed photovoltaic power transmission, quantifying the risk of regional voltage instability. The impact indicators corresponding to power flow security risks may include: Line load factor is used to measure the extent to which the predicted current of the line exceeds the rated current after distributed photovoltaic (PV) grid connection. Reverse load factor measures the proportion of power fed back to the grid by distributed photovoltaic power, reflecting the degree of grid power flow reversal. Impact indicators corresponding to frequency security risks may include: Frequency volatility is a key indicator used to measure the speed and magnitude of changes in power system frequency over a short period of time. It reflects the dynamic stability of the system frequency and helps to prevent excessive frequency deviations caused by distributed photovoltaic (PV) integration, which could lead to equipment shutdowns or system crashes. Frequency exceedance rate is an indicator used to measure the proportion of time or events in which the frequency of a power system exceeds the safety limit, reflecting the risk of system frequency stability. The frequency response reserve factor, which is the ratio of the system's reserve capacity to the maximum load, is used to reflect the ability to ensure rapid frequency recovery and prevent low-frequency load shedding during sudden load changes or failures. The above indicators constitute the distributed photovoltaic power transmission safety early warning indicator system, as shown in Table 1: Table 1. Distributed Photovoltaic Power Transmission Safety Early Warning Indicator System

[0027] Extract indicator features based on each influencing indicator; Establish a second mapping relationship between the aforementioned indicator characteristics and the corresponding risk application scenarios; The determination of voltage security risk, power flow security risk, and frequency security risk after grid integration of distributed photovoltaic power generation based on the distributed photovoltaic power output forecast and load demand forecast includes: Based on the predicted output and load demand of the distributed photovoltaic power generation, determine the respective impact index values ​​of voltage security risk, power flow security risk and frequency security risk after the grid is connected to distributed photovoltaic power generation. The schematic diagram of the constructed distributed photovoltaic grid connection is shown below. Figure 2 As shown in the figure, the calculation method for each influencing indicator value in this embodiment is as follows: Node voltage deviation rate:

[0028]

[0029] In the formula, U 预测 U represents the predicted node voltage. 额定 R0 represents the rated voltage, Li is the length of the power supply line, L1 represents the distance between the distributed photovoltaic grid connection point and the outlet switch of the upstream substation, L2 represents the distance between the distributed photovoltaic grid connection point and the load, and R0 represents the line impedance.

[0030] Node voltage fluctuation:

[0031] Node voltage over-limit rate:

[0032] In the formula, Np and N represent the number of node voltage over-limits and the total number of measurements, respectively. Line load rate:

[0033] In the formula, I 预测 I 额定 These are the predicted and rated values ​​of the line current, U. j Indicates the voltage of the corresponding branch; Reverse load rate:

[0034] In the formula, P 倒送 P and P represent the predicted power of distributed photovoltaic backfeed and the total power of the grid, respectively. Frequency volatility:

[0035] In the formula, K fP is the power grid available frequency regulation coefficient. G Power supplied to the power grid; Frequency exceeding limit rate:

[0036] In the formula, M f M and M represent the number of times the power grid frequency exceeded the limit and the total number of measurements, respectively. Frequency response reserve factor:

[0037] In the formula, P represents the reserve coefficient. 备用 P L,max (d, t) represent the system's reserve capacity and the predicted maximum load, respectively. The risk application scenarios based on the matching of voltage security risk, power flow security risk, and frequency security risk include: The corresponding indicator characteristics are determined based on the impact index values ​​of voltage safety risk, power flow safety risk, and frequency safety risk, respectively. Based on the characteristics of the indicators and the second mapping relationship, the corresponding risk application scenarios are matched.

[0038] In another alternative implementation, the indicator feature includes indicator duration; Specifically, the predicted voltage safety, power flow safety, and frequency safety indicators at each time point are compared with their corresponding over-limit thresholds. By accumulating the time intervals between over-limit moments, the cumulative over-limit time for each indicator can be obtained.

[0039] In the formula, T represents the cumulative time exceeding the limit, and t k This represents the time limit exceeded for the kth time. This is a unit of time statistics.

[0040] The establishment of a second mapping relationship between the indicator features and the corresponding risk application scenarios includes: If the duration of the indicators corresponding to three different types of security risks exceeds the first preset threshold within a first preset time, then the corresponding risk application scenario is the first risk application scenario: the power grid faces a direct threat. If the duration of the indicators corresponding to two different types of safety risks exceeds the first preset threshold within a second preset time, then the corresponding risk application scenario is the second risk application scenario: the local power grid operation exceeds the safety margin. If there is an indicator whose duration exceeds the second preset threshold but is less than the first preset threshold within the third preset time, the corresponding risk application scenario is the third risk application scenario: the indicator deviates from the normal range, but does not threaten the power grid safety for the time being. The warning level corresponding to the first risk application scenario is Level 1 warning; The warning level corresponding to the second risk application scenario is Level II warning; The warning level corresponding to the third risk application scenario is Level 3 warning. In practice, to prevent cascading failures such as voltage collapse and frequency instability, protect equipment from overload damage, dynamically optimize source-load balance, and minimize economic losses from power outages, this implementation method is based on the distributed photovoltaic power transmission safety early warning index system. It proposes early warning judgment rules based on the dual dimensions of "severity + duration", realizing closed-loop management from instantaneous impact early warning to cumulative risk prevention and control, making up for the lag in traditional post-event handling, and is especially suitable for areas where the distributed photovoltaic carrying capacity assessment level is red. The principles for classifying warning levels are as follows: Level 1 Warning (Highest Warning): The power grid safety is under direct threat, which may trigger a cascading failure or a large-scale power outage. Multiple nodes may experience simultaneous voltage exceedances, and critical equipment may be severely overloaded, triggering the warning within a short time. At Level 1 warning, an immediate response and automatic implementation of protective measures are required.

[0041] Level 2 Warning (Intermediate Warning): Local power grid operation exceeds safety margins, requiring timely intervention. This can be triggered within 10-20 minutes by single-node voltage exceeding limits, continuous line overload, and increased frequency fluctuations. At Level 2 warning, a rapid response and automatic adjustment of equipment parameters are necessary.

[0042] Level 3 warning (basic warning): The indicators deviate from the normal range but do not pose a complete threat to the power grid. The voltage / frequency deviation is close to the limit and the equipment load rate is too high. If this continues for about 30 minutes, a safety warning will be triggered. After the warning notice is issued, monitoring needs to be strengthened. For example, if at least three different types of indicators exceed their corresponding severity thresholds within a short period of time (<10 minutes), a Level 1 warning can be triggered. For instance, if the node voltage over-limit rate exceeds the severe over-limit ratio (20%), the line load rate exceeds the severe overload ratio (120%), and the frequency over-limit rate exceeds the severe over-limit percentage (20%), then a Level 1 warning will be triggered.

[0043] There are two different types of indicators. If they exceed the threshold corresponding to their severity within 10 to 20 minutes, a level 2 warning will be triggered. For example, if the node voltage over-limit rate exceeds a certain over-limit ratio (15%) and the reverse heavy overload remains at a high level (20% to 30%), a level 2 warning will be triggered.

[0044] A Level 3 warning is triggered when an indicator deviates from the normal range for approximately 30 minutes, approaching but not reaching the threshold for severe or moderate severity. For example, a Level 3 warning is triggered if the voltage deviation rate approaches the allowable normal range boundary (±5%), or if the line load rate is high and close to the rated load rate limit, etc., for approximately 30 minutes. The final classification model for the safety early warning level of distributed photovoltaic power transmission is shown in Table 2: Table 2 Classification Model of Distributed Photovoltaic Power Transmission Safety Early Warning Levels

[0045] In another alternative implementation, it further includes: Decisions are made and corresponding instructions are issued based on the warning level. In practice, if a Level 1 warning is triggered, some non-critical distributed photovoltaic power sources should be immediately cut off, and photovoltaic equipment with larger power and less impact on power supply should be selected first; backup lines should be started to reduce the power transmission of this line; and large-scale frequency regulation power sources (energy storage power stations or traditional power sources with frequency regulation capabilities) should be started quickly to adjust the system frequency rapidly.

[0046] If a Level 2 warning is triggered, the distributed photovoltaic power output needs to be fine-tuned and the voltage needs to be initially adjusted; an instruction should be issued for the distributed photovoltaic power station to appropriately reduce the power output, adjusting the output power to a reasonable range close to the rated load; some small frequency regulation devices (distributed energy storage devices) can be appropriately connected to adjust the frequency.

[0047] If a Level 3 warning is triggered, the encrypted monitoring mode of the voltage monitoring system needs to be activated. Every preset time, such as 10 minutes, the node voltage data should be recorded to closely track voltage changes. The line load should be closely monitored, and preparations should be made to adjust the power at any time. The frequency monitoring points should be increased and the frequency monitoring should be encrypted.

[0048] Once a decision is made, instructions are precisely transmitted to distributed photovoltaic power stations, operation and maintenance personnel, and other relevant departments through intelligent control systems and communication networks. For distributed photovoltaic power stations, instructions are transmitted to the station's monitoring center, where the monitoring system automatically executes operations such as adjusting inverter power and changing reactive power output. For operation and maintenance personnel, detailed on-site operation instructions are sent via SMS, mobile applications, etc., clearly defining work tasks and timelines to ensure rapid response and execution. For grid dispatch departments and other relevant units, instructions are transmitted through dispatch communication systems to facilitate collaboration and joint response to early warning situations.

[0049] During the instruction issuance process, a feedback mechanism is set up to confirm the receipt and execution of instructions in real time, ensuring that decisions can be implemented efficiently.

[0050] In another alternative implementation, such as Figure 3 As shown, a distributed photovoltaic power transmission risk early warning device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of a distributed photovoltaic power transmission risk early warning method described in any of the above embodiments.

[0051] In practical implementation, a data layer, a probability prediction layer, a model building layer, and an early warning decision layer can be constructed. Among them, the data layer is responsible for collecting and processing various types of data; the probability prediction layer uses historical data to predict the operating results of various key parameters; the model building layer calculates early warning-related indicators based on data; and the early warning decision layer triggers early warnings based on the calculation results and preset rules.

[0052] In another alternative embodiment, a computer-readable storage medium stores computer program instructions thereon, which, when executed by a processor, implement the steps of a distributed photovoltaic power transmission risk warning method as described in any of the above embodiments.

[0053] In summary, the distributed photovoltaic (PV) power transmission risk early warning method, device, and computer-readable storage medium provided by this invention effectively fill the gaps in traditional research on risk assessment systems, which lack systematicity, comprehensiveness, and quantifiability, as well as multi-factor coupled risk early warning methods. It clarifies the quantitative indicators and grading standards for distributed PV power transmission risks. Addressing the grid transmission security risks arising from increased distributed PV penetration, this invention, unlike traditional research, comprehensively considers factors such as voltage, power, frequency, stability, power quality, and reliability to construct a systematic, comprehensive, and quantifiable risk assessment model. This model accurately identifies and assesses risks during distributed PV power transmission, improving the accuracy and reliability of risk early warning, providing more precise decision-making basis for grid dispatching and operation, and better addressing the challenges brought by distributed PV grid integration. (1) Based on multiple factors such as stability, power quality, and reliability, construct a quantitative index system for distributed photovoltaic power transmission safety risks covering dimensions such as voltage safety, power flow safety, and frequency safety, clarify risk classification standards, and meet the needs of refined grid dispatching; (2) Unlike traditional research that focuses on isolated analysis of a single factor, this invention fully considers the coupled impact of multiple factors such as meteorological conditions, power grid topology, and energy storage system configuration on the risk of distributed photovoltaic power transmission. Through the coordinated operation of the probability prediction layer, model building layer, and early warning decision layer, it achieves accurate early warning of the risk of multiple coupled factors. (3) Breaking through the traditional single-indicator early warning model, an innovative dual-dimensional early warning judgment model based on "severity + duration" is proposed. By quantitatively analyzing the degree of exceedance and duration of various indicators such as voltage, power flow, and frequency, the first, second, and third level early warning levels are clearly defined. A dynamic response mechanism covering the entire process of decision-making, instruction issuance, and execution feedback is established. Differentiated control strategies, including power supply switching, power regulation, and encrypted monitoring, are formulated for different early warning levels. (4) In the probability prediction layer, historical data and statistical analysis methods are combined to construct a distributed photovoltaic power output prediction model and a load demand prediction model. Through per-unit value calculation and statistical model, the future distributed photovoltaic power output and load demand are accurately predicted, providing dynamic data support for risk warning.

[0054] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent modifications made based on the content of the present invention specification and drawings, or direct or indirect applications in related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A distributed photovoltaic power generation risk early warning method, characterized in that, The method comprises the following steps: collecting power grid operation data and distributed photovoltaic data, determining distributed photovoltaic output prediction value and load demand prediction value according to the power grid operation data and the distributed photovoltaic data; determining voltage safety risk, power flow safety risk and frequency safety risk of the power grid after connecting the distributed photovoltaic according to the distributed photovoltaic output prediction value and the load demand prediction value; matching corresponding risk application scenarios based on the voltage safety risk, the power flow safety risk and the frequency safety risk, and determining corresponding warning levels according to the risk application scenarios. 2.The distributed photovoltaic risk early warning method according to claim 1, characterized in that, It also includes collecting meteorological data; the step of determining distributed photovoltaic output prediction value and load demand prediction value according to the power grid operation data and the distributed photovoltaic data comprises: determining load demand prediction value according to the power grid operation data; determining distributed photovoltaic output prediction value according to the distributed photovoltaic data and the meteorological data. 3.The distributed photovoltaic risk early warning method according to claim 1, characterized in that, After collecting the power grid operation data and the distributed photovoltaic data, it further comprises the following steps: for the collected data, it is determined whether it is out of the preset range by 3σ principle, if yes, the data out of the preset range is removed, and the linear interpolation method is used to fill in the vacancy.

4. The distributed photovoltaic risk early warning method according to any one of claims 1 to 3, characterized in that, It also includes: determining a first mapping relationship between risk application scenarios and warning levels in advance; the step of determining corresponding warning levels according to the risk application scenarios comprises: determining corresponding warning levels according to the risk application scenarios and the first mapping relationship.

5. The distributed photovoltaic risk early warning method according to any one of claims 1 to 3, characterized in that, It also includes: determining impact indicators corresponding to the voltage safety risk, the power flow safety risk and the frequency safety risk respectively in advance; extracting index characteristics according to each impact indicator; establishing a second mapping relationship between the index characteristics and the corresponding risk application scenarios; the step of determining voltage safety risk, power flow safety risk and frequency safety risk of the power grid after connecting the distributed photovoltaic according to the distributed photovoltaic output prediction value and the load demand prediction value comprises: determining impact indicator values corresponding to the voltage safety risk, the power flow safety risk and the frequency safety risk respectively of the power grid after connecting the distributed photovoltaic according to the distributed photovoltaic output prediction value and the load demand prediction value; the step of matching corresponding risk application scenarios based on the voltage safety risk, the power flow safety risk and the frequency safety risk comprises: determining corresponding index characteristics according to the impact indicator values corresponding to the voltage safety risk, the power flow safety risk and the frequency safety risk respectively; matching corresponding risk application scenarios according to the index characteristics and the second mapping relationship.

6. The distributed photovoltaic risk early warning method according to claim 5, characterized in that, The index characteristics include index duration; the step of establishing a second mapping relationship between the index characteristics and the corresponding risk application scenarios comprises: if the index duration corresponding to three different types of safety risks exceeds a first preset threshold within a first preset time, the corresponding risk application scenario is a first risk application scenario: the power grid is directly threatened; if the index duration corresponding to two different types of safety risks exceeds the first preset threshold within a second preset time, the corresponding risk application scenario is a second risk application scenario: the local power grid operation exceeds the safety margin; If there is an index that exceeds the second preset threshold and is less than the first preset threshold for a third preset time, the corresponding risk application scenario is a third risk application scenario: the index deviates from the normal range, but does not threaten the power grid safety.

7. The distributed photovoltaic risk early warning method according to claim 6, characterized in that, The first risk application scenario corresponds to a first warning level; The second risk application scenario corresponds to a second warning level; The third risk application scenario corresponds to a third warning level.

8. The distributed photovoltaic risk early warning method according to any one of claims 1 to 3, characterized in that, Further comprising: According to the warning level, a corresponding decision is made, and a corresponding instruction is issued.

9. A distributed photovoltaic risk early warning device, comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the steps of the distributed photovoltaic risk early warning method according to any one of claims 1-8.

10. A computer-readable storage medium having stored thereon computer program instructions, wherein, The computer program instructions are executed by the processor to implement the steps of the distributed photovoltaic risk early warning method according to any one of claims 1-8.

Images