Novel power distribution system multi-scene risk situation assessment method

Abstract

The invention provides a novel power distribution system multi-scene risk situation assessment method. The method comprises the following steps: firstly, constructing a wind / light / load joint uncertainty model, and generating two typical scenes of conventional operation and N-k emergency by utilizing probabilistic power flow simulation; establishing a multi-dimensional risk index system considering voltage, thermal stability and power flow direction, and covering indexes such as voltage out-of-limit, voltage deviation, overload, load deviation, reverse power flow, N-1 passing rate and the like; data objective distribution characteristics are mined by adopting an entropy weight method, subjective experience of experts is reflected in combination with an analytic hierarchy process, and weights of all indexes are determined by constructing a subjective and objective combination weighting model; and calculating a comprehensive risk assessment value of the power distribution system and dividing a'safety-early warning-danger 'multi-level risk situation. According to the method, the small probability and high risk under high-load, high-permeability, conventional operation and N-k emergency complex scenes can be effectively identified, and quantitative support is provided for power distribution system operation mode optimization and weak link treatment.

Description

Technical Field

[0001] This invention belongs to the field of power system operation and dispatch technology, and relates to a power system risk assessment method, and more specifically, to a novel multi-scenario risk situation assessment method for distribution systems. Background Technology

[0002] With the construction of new power systems, the penetration rate of distributed energy in distribution networks is constantly increasing. Distribution networks containing large-scale distributed energy resources face significant risk exposure due to the dual uncertainties of source and load, multiple concurrent faults, and frequent switching of operating modes.

[0003] Existing risk assessment methods for distribution networks have developed a relatively comprehensive indicator system. Commonly used indicators focus on voltage deviation, branch power flow exceeding limits, and load shedding. For high-penetration photovoltaic scenarios, some studies have constructed system-level voltage exceeding limit and branch overload risk indices based on probabilistic power flow; for reverse power flow problems, some studies have characterized the risk of reverse heavy overload at multiple levels from "transformer area - feeder - station". At the assessment method level, researchers have proposed diverse models based on cloud models, graph neural networks, distributed entropy, and Telehl entropy.

[0004] However, while the existing theoretical framework is relatively mature, some technical problems still exist: existing methods are mostly limited to static current flow or a small number of scenarios, making it difficult to take into account both routine operation and emergency scenarios, resulting in insufficient assessment of local risks under high penetration rates; existing weighting methods either rely solely on objective data, ignoring practical engineering significance, or rely solely on subjective judgment, lacking data support. Differences in risk tolerance among different regions and decision-makers are not systematically reflected; traditional methods focus excessively on high-probability routine scenarios, making it difficult to account for low-probability but serious failure events, and the interpretability of decisions is weak.

[0005] Therefore, there is an urgent need for an assessment method that can integrate subjective and objective information, cover complex operating scenarios, and accurately depict risk situations, so as to provide quantitative support for optimizing the operation mode of the power distribution network and addressing its weaknesses. Summary of the Invention

[0006] This invention provides a novel multi-scenario risk situation assessment method for power distribution systems, aiming to solve the problems of imprecise risk characterization of power distribution networks in complex scenarios, difficulty in balancing subjective and objective weights, and weak ability to identify risks with low probability but high consequences.

[0007] This invention provides a novel multi-scenario risk assessment method for power distribution systems, comprising the following steps:

[0008] 1) Determine the network topology and the location of distributed energy access, establish a source-load uncertainty probability model that includes the output and load of distributed energy sources such as wind power and photovoltaics, and consider the correlation of distributed energy output;

[0009] 2) Utilize Latin hypercube sampling to generate large-scale input samples, perform probabilistic power flow calculations, and construct a set of regular operation scenarios covering different load levels and distributed energy penetration rates, as well as an Nk emergency scenario set containing the failure of k components;

[0010] 3) For routine operation scenarios, establish indicators for node voltage over-limit risk, branch line overload risk, and power backfeed risk based on a frequency-severity framework; for Nk emergency scenarios, establish indicators for node voltage deviation risk, branch line load change risk, and N-1 pass level.

[0011] 4) Construct an objective weighting model based on the entropy weighting method and a subjective weighting model based on the analytic hierarchy process, calculate the objective weights and subjective weights of each risk indicator, and obtain the combined weights by merging them through the linear weighted normalization method.

[0012] 5) Combining the distribution and combined weights of various risk indicators for complex operating scenarios obtained in step 3), calculate the risk status assessment value of the power distribution system, and determine the risk level of the power distribution network according to the preset risk classification standard.

[0013] In step 1) of this invention, the establishment of the source-load uncertainty probability model specifically includes: using the Weibull distribution to characterize the wind speed probability model, and calculating the output power of the wind turbine through a three-segment cubic curve of the cut-in wind speed, rated wind speed, and cut-out wind speed; using the Beta distribution to characterize the light intensity probability model, and calculating the photovoltaic output power based on the light intensity, photovoltaic array area, and conversion efficiency.

[0014] In step 2) of this invention, a truncated normal distribution with adjustable mean and variance is used to characterize the active and reactive power of the base load, and the correlation between wind, solar and load is handled by introducing a covariance matrix and Cholesky decomposition.

[0015] In step 3) of this invention, the total number of time sections is N. t The method for calculating risk indicators for typical operational scenarios at each time segment is as follows:

[0016] Node voltage exceedance risk: calculated based on the degree of deviation of node voltage from set upper and lower limits.

[0017]

[0018] Where: R VD,i V represents the voltage deviation risk in the i-th time period; j,k For system state Ek Voltage value at node j; V max V min These are the set upper and lower bounds of the system voltage; ΔV max The maximum acceptable voltage deviation; N b This represents the total number of nodes.

[0019] Tributary line overload risk: calculated based on the degree to which the tributary load rate exceeds the maximum allowable transmission load rate.

[0020]

[0021] Among them, R OL,i For the line overload risk in the i-th time period; l j,k For system state E k Load rate of line j; l j,max N represents the maximum load rate allowed for transmission on branch j. L This represents the total number of branch roads.

[0022] Power backfeed risk: calculated based on the percentage of reverse power at the starting line relative to the line's operating limit.

[0023]

[0024] Among them, R PR,i For the power backfeed risk in the i-th time period; l PR,k For system state E k The percentage of the power of the downstream line relative to the operating limit of the power grid transformer or line; if reverse transmission occurs, the power is a negative value.

[0025] The method for calculating the risk index of emergency scenario Nk at each time segment is as follows:

[0026] Node voltage deviation risk: Using the baseline voltage before the fault as a reference, calculate the deviation of the node voltage after the fault and map it to severity.

[0027]

[0028] Wherein, the voltage at the node after disconnection is V. j,k V0 represents the node voltage value before the system disturbance, R j,k It is determined by the degree to which the node voltage exceeds the limit before the disturbance.

[0029] Branch load change risk: Calculate the degree of change in branch load rate after a fault relative to the load rate before the fault.

[0030]

[0031] Among them, the load rate after disconnection is l j,k Before the system disturbance, the branch load rate was l0, Rj,k It is determined by the degree of overload of the branch before the disturbance.

[0032] N-1 pass rate: Statistics on the proportion of system states that meet the requirements for node voltage and line load rate in all generated emergency combinations, mainly including line N-1 achievement rate and node N-1 achievement rate.

[0033]

[0034] Among them, R LC,i Let l be the route N-1 completion rate in the i-th time period; LC,k For state E k The following line ratio satisfies N-1, B LC,k For state E k The following satisfies the node ratio of N-1.

[0035] In step 4) of this invention, the construction steps of the objective weighting model based on the entropy weight method include:

[0036] Construct the probability distribution matrix of risk indicators: Calculate the probability that the j-th indicator falls within the level range of the i-th indicator.

[0037]

[0038] Where, x ij This represents the probability that the j-th indicator falls within the level range of the i-th indicator.

[0039] The information entropy of each indicator is calculated based on the relative frequency of the indicator data. The smaller the information entropy, the greater the variability of the indicator.

[0040]

[0041] Objective weights are calculated based on information entropy, and the objective weights of indicators are inversely proportional to their information entropy:

[0042]

[0043] Among them, H j Let be the entropy weight of the j-th index; 1-E j This represents the degree of variability corresponding to the j-th indicator. The larger the information entropy, the smaller its degree of variability, and the smaller its impact on the overall risk assessment of the power distribution system.

[0044] The construction steps of the subjective weighting model based on the analytic hierarchy process are as follows:

[0045] First, a decision matrix P is established, using a scale of 1 to 9 to measure the relative importance between any two risk indicators. Then, the largest eigenvalue of the decision matrix and its corresponding eigenvector are calculated. Finally, matrix P is normalized to obtain the relative weights W of the decision indicators using the analytic hierarchy process (AHP). j .

[0046]

[0047] Where, λ max represents the largest eigenvalue of the decision matrix; n is the order of the decision matrix. When the consistency ratio CR is less than 0.1, it indicates that the overall consistency of the hierarchical ranking results under subjective thinking is relatively high. RI is the random consistency index for the corresponding n-dimensional matrix, and RI increases with the increase of the matrix order. A consistency check is performed on the decision matrix; when the consistency ratio is less than 0.1, the eigenvectors are normalized and used as subjective weights.

[0048] After obtaining the weights of the subjective and objective indicators, the combined weights of each indicator are obtained using the following method:

[0049]

[0050] Where n is the number of indicators; H j W j These represent the subjective and objective weights of the j-th indicator, respectively. The combined weight coefficients are calculated, then normalized. Finally, the final weight value is obtained by linearly calculating the coefficients with the subjective and objective weights.

[0051] The specific process of step 5) of this invention is as follows: multiply the risk impact coefficient, the normalized probability index and the combined weight, and use a weighted summation method to obtain the distribution network risk situation assessment value E under complex risk scenarios.

[0052]

[0053] Among them, a ij x represents the impact coefficient of different risk levels on the power distribution system. ij S is a normalized probability index. j The weights are then combined. The risk status of the power distribution system is then classified into three levels: safe, warning, and dangerous.

[0054] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0055] Previous risk assessment methods for distribution networks have largely been limited to static power flow or a few routine scenarios, making it difficult to consider both routine operation and emergency scenarios. Furthermore, they often focus excessively on high-probability events, resulting in insufficient characterization of localized penetration risks under high penetration conditions. This leads to overly optimistic assessments of risk levels in typical operating conditions such as high load and high penetration. The multi-scenario risk assessment method for distribution networks with high-penetration distributed generation provided in this invention constructs a multi-dimensional risk indicator system for both routine operation and emergency scenarios, mapping it uniformly into a frequency-severity framework. This effectively exposes tail risks with low probability but high consequences in emergency scenarios, demonstrating greater sensitivity in identifying localized high risks under high load and high penetration conditions, and providing operators with more room for adjustment. Previous assessment indicator weighting mechanisms were often simplistic. Relying solely on objective weighting methods such as entropy weighting easily overlooks practical engineering significance and safety preferences, while relying solely on subjective weighting methods such as the analytic hierarchy process lacks objective data support and fails to systematically reflect the differences in risk acceptability among different decision-makers. The proposed subjective-objective combined weighting model combines the data mining capabilities of the entropy weighting method with the expert experience of the analytic hierarchy process (AHP). While inheriting experts' engineering preferences for key safety indicators, it introduces objective data-driven information, resulting in comprehensive evaluation results that are both engineering interpretable and possess higher discriminative power and safety margin. Furthermore, previous risk status assessments often employed a static "safe / unsafe" binary classification, which is insufficient to meet the needs of refined management in modern power systems. This invention, based on the combined weighting model, calculates risk status assessment values, enabling dynamic monitoring and hierarchical classification of multi-level risk statuses in distribution networks ("safe—early warning—dangerous"). It can intuitively reflect the impact of complex operating scenarios on distribution network safety, providing quantitative scientific basis for optimizing operating modes, addressing weak links, and making early warning decisions under different load levels and DG penetration rates. Attached Figure Description

[0056] Figure 1 This is a schematic diagram of the method flow of the present invention.

[0057] Figure 2 This is the system architecture diagram of the IEEE-33 node after the numbering. Detailed Implementation

[0058] The technical solution of the present invention will be described in detail below with reference to the embodiments and the accompanying drawings.

[0059] Figure 1 This is a schematic diagram of the method flow of the present invention, which introduces the basic steps of the method of the present invention. Figure 2 This refers to the IEEE-33 node system, which is numbered. The following description uses this system as an example to illustrate the specific implementation of the method of this invention.

[0060] 1) The system rated voltage is 12.66kV. Wind turbines with a rated power of 1200kW are connected at nodes 9 and 17; photovoltaic (PV) systems with a capacity of 600kW are connected at nodes 4 and 32. A Weibull distribution is used to characterize the wind speed probability model, and the output power of the wind turbines is calculated using a three-segment cubic curve of the cut-in wind speed, rated wind speed, and cut-out wind speed. A Beta distribution is used to characterize the irradiance probability model, and the PV output power is calculated based on irradiance, PV array area, and conversion efficiency.

[0061] 2) Based on step 1), since the load follows a normal distribution, the load injection for the i-th scenario bus n is generated using the "mean × random scaling" method, with the sampling number N set. s For 2000 iterations, a random input vector is generated, and probabilistic power flow calculations are performed using Latin hypercube sampling to obtain the distribution of node voltage and branch power.

[0062] 3) Building upon step 2), to characterize the multi-dimensional safety attributes of the power distribution system under normal operation and emergency conditions, this invention establishes a hierarchical index level based on the probabilistic power flow results. This establishes the index basis for subsequent risk assessment and classification of the power distribution system, as shown in the table below:

[0063] Indicator Level Voltage over-limit Line overload Power backfeed Voltage deviation Load changes N-1 pass rate Low [0,0.03] [0,0.15] 0 [0,0.2] [0,0.2] 1 lower [0.03,0.05] [0.15,0.3] [0,0.1] [0.2,0.4] [0.2,0.4] [0.9,1] medium [0.05,0.07] [0.3,45] [0.1,0.2] [0.4,0.6] [0.4,0.6] [0.8,0.9] higher [0.07,0.09] [0.45,0.6] [0.2,0.3] [0.6,0.8] [0.6,0.8] [0.7,0.8] high [0.09,+∞] [0.6,+∞] [0.3,+∞] [0.8,+∞] [0.8,+∞] [0,0.7]

[0064] The total number of time sections is N t The total number of time segments for the conventional operation scenarios in this invention is 2000. The method for calculating the risk index of the conventional operation scenario under each time segment is as follows:

[0065] Node voltage exceedance risk: calculated based on the degree of deviation of node voltage from set upper and lower limits.

[0066]

[0067] Where: R VD,i V represents the voltage deviation risk in the i-th time period; j,k For system state E k Voltage value at node j; V max V min These are the set upper and lower bounds of the system voltage; ΔV max The maximum acceptable voltage deviation; N b This represents the total number of nodes.

[0068] Tributary line overload risk: calculated based on the degree to which the tributary load rate exceeds the maximum allowable transmission load rate.

[0069]

[0070] Among them, R OL,iFor the line overload risk in the i-th time period; l j,k For system state E k Load rate of line j; l j,max N represents the maximum load rate allowed for transmission on branch j. L This represents the total number of branch roads.

[0071] Power backfeed risk: calculated based on the percentage of reverse power at the starting line relative to the line's operating limit.

[0072]

[0073]

[0074] Among them, R PR,i For the power backfeed risk in the i-th time period; l PR,k For system state E k The percentage of the power of the downstream line relative to the operating limit of the power grid transformer or line; if reverse transmission occurs, the power is a negative value.

[0075] The total number of time segments for the Nk emergency scenarios in this invention is 128,000. The method for calculating the risk index of the Nk emergency scenarios under each time segment is as follows:

[0076] Node voltage deviation risk: Using the baseline voltage before the fault as a reference, calculate the deviation of the node voltage after the fault and map it to severity.

[0077]

[0078] Wherein, the voltage at the node after disconnection is V. j,k V0 represents the node voltage value before the system disturbance, R j,k It is determined by the degree to which the node voltage exceeds the limit before the disturbance.

[0079] Branch load change risk: Calculate the degree of change in branch load rate after a fault relative to the load rate before the fault.

[0080]

[0081] Among them, the load rate after disconnection is l j,k Before the system disturbance, the branch load rate was l0, R j,k It is determined by the degree of overload of the branch before the disturbance.

[0082] N-1 pass rate: Statistics on the proportion of system states that meet the requirements for node voltage and line load rate in all generated emergency combinations, mainly including line N-1 achievement rate and node N-1 achievement rate.

[0083]

[0084] Among them, R LC,i Let l be the route N-1 completion rate in the i-th time period; LC,k For state E k The following line ratio satisfies N-1, B LC,k For state E k The following satisfies the node ratio of N-1.

[0085] 4) Based on step 3), the influence of each indicator on the overall objective is not the same. Simply weighting them equally may mask indicators more sensitive to accident tendencies, leading to biases in equipment vulnerability identification, zoned management, and resource allocation. To balance the objective characteristic distribution given by the sample data with expert knowledge's judgment of the project's importance, this invention proposes a combined weighting method integrating objective and subjective weighting for the final calculation and classification of the comprehensive risk situation of the power distribution system.

[0086] The steps for constructing an objective weighting model based on the entropy weight method include:

[0087] Construct the probability distribution matrix of risk indicators: Calculate the probability that the j-th indicator falls within the level range of the i-th indicator.

[0088]

[0089] Where, x ij This represents the probability that the j-th indicator falls within the level range of the i-th indicator. This invention employs a total of 6 risk indicators and 5 indicator levels. To avoid numerical singularities in logarithmic operations, when xij = 0, a very small amount of 10 is used. -10 It replaces, but does not change the overall trend.

[0090] The information entropy of each indicator is calculated based on the relative frequency of the indicator data. The smaller the information entropy, the greater the variability of the indicator.

[0091]

[0092] Objective weights are calculated based on information entropy, and the objective weights of indicators are inversely proportional to their information entropy:

[0093]

[0094] Among them, H j Let be the entropy weight of the j-th index; 1-E j This represents the degree of variability corresponding to the j-th indicator. The larger the information entropy, the smaller its degree of variability, and the smaller its impact on the overall risk assessment of the power distribution system.

[0095] The construction steps of the subjective weighting model based on the analytic hierarchy process are as follows:

[0096] First, a decision matrix P is established, using a scale of 1 to 9 to measure the relative importance between any two risk indicators, as shown in the table below. Then, the maximum eigenvalue λmax of the decision matrix P and its corresponding eigenvector are calculated. Finally, matrix P is normalized to obtain the relative weights W of the decision indicators using the analytic hierarchy process (AHP). j .

[0097] Scale meaning 1 a and b are equally important 3 a is slightly more important than b. 5 A is significantly more important than B. 7 a is more important than b 9 a is extremely important than b.

[0098] When constructing the decision matrix, subjective judgments regarding the overall importance of indicators may contain significant errors. Therefore, a consistency check is performed on the decision matrix P. The consistency ratio of the decision matrix P and the formulas for calculating the indicators are as follows:

[0099]

[0100] Where, λ max denoted by , where is the largest eigenvalue of the decision matrix; and 'n' is the order of the decision matrix. When the consistency ratio CR is less than 0.1, it indicates that the overall consistency of the hierarchical ranking results under subjective thinking is relatively high. RI is the random consistency index for the corresponding n-dimensional matrix, and RI increases with the increase of the matrix order, as shown in the table below.

[0101] Indicator / Dimension n 1 2 3 4 5 6 RI 0.00 0.00 0.58 0.94 1.12 1.24

[0102] The decision matrix is ​​checked for consistency. When the consistency ratio is less than 0.1, the eigenvectors are normalized and used as subjective weights.

[0103] After obtaining the weights of the subjective and objective indicators, the combined weights of each indicator are obtained using the following method:

[0104]

[0105] Where n is the number of indicators; H j W j These represent the subjective and objective weights of the j-th indicator, respectively. The combined weight coefficients are calculated, then normalized. Finally, the final weight value is obtained by linearly calculating the coefficients with the subjective and objective weights.

[0106] 5) Based on step 4), the risk impact coefficient, normalized probability index and combined weight are multiplied together, and the weighted summation method is used to obtain the distribution network risk status assessment value E under complex risk scenarios.

[0107]

[0108] Among them, a ij x represents the impact coefficient of different risk levels on the power distribution system. ij S is a normalized probability index.j The weights are then combined. The risk status of the power distribution system is then divided into three levels: safe, warning, and dangerous, as shown in the table below.

[0109]

[0110]

[0111] The above are merely preferred embodiments of the present invention. It should be noted that for those skilled in the art, several foreseeable improvements and equivalent substitutions can be made without departing from the principle of the present invention. All such technical solutions after improvements and equivalent substitutions to the claims of the present invention fall within the protection scope of the present invention.

Claims

1. A novel multi-scenario risk assessment method for power distribution systems, characterized in that, The method includes the following steps: 1) Determine the network topology and the location of distributed energy access, establish a source-load uncertainty probability model that includes the output and load of distributed energy sources such as wind power and photovoltaics, and consider the correlation of distributed energy output; 2) Utilize Latin hypercube sampling to generate large-scale input samples, perform probabilistic power flow calculations, and construct a set of regular operation scenarios covering different load levels and distributed energy penetration rates, as well as an Nk emergency scenario set containing the failure of k components; 3) For routine operation scenarios, establish indicators for node voltage over-limit risk, branch line overload risk, and power backfeed risk based on a frequency-severity framework; for Nk emergency scenarios, establish indicators for node voltage deviation risk, branch line load change risk, and N-1 pass level. 4) Construct an objective weighting model based on the entropy weighting method and a subjective weighting model based on the analytic hierarchy process, calculate the objective weights and subjective weights of each risk indicator, and obtain the combined weights by merging them through the linear weighted normalization method. 5) Combining the distribution and combined weights of various risk indicators for complex operating scenarios obtained in step 3), calculate the risk status assessment value of the power distribution system, and determine the risk level of the power distribution network according to the preset risk classification standard.

2. The multi-scenario risk situation assessment method for a novel power distribution system according to claim 1, wherein step 1) of establishing the source-load uncertainty probability model specifically includes: The Weibull distribution is used to characterize the wind speed probability model, and the output power of the wind turbine is calculated by a three-segment cubic curve of the cut-in wind speed, rated wind speed, and cut-out wind speed. The Beta distribution is used to characterize the light intensity probability model, and the photovoltaic output power is calculated based on the light intensity, photovoltaic array area, and conversion efficiency.

3. In the multi-scenario risk situation assessment method for the novel power distribution system according to claim 1, in step 2), a truncated normal distribution with adjustable mean and variance is used to characterize the active and reactive power of the base load, and the correlation between wind, solar and load is processed by introducing a covariance matrix and Cholesky decomposition.

4. In the multi-scenario risk assessment method for a novel power distribution system according to claim 1, in step 3), the total number of time segments is N. t The method for calculating risk indicators for typical operational scenarios at each time segment is as follows: Node voltage exceedance risk: calculated based on the degree of deviation of node voltage from set upper and lower limits. in: R VD,i V represents the voltage deviation risk in the i-th time period; j,k For system state E k Voltage value at node j; V max V min These are the set upper and lower bounds of the system voltage; ΔV max The maximum acceptable voltage deviation; N b This represents the total number of nodes. Tributary line overload risk: calculated based on the degree to which the tributary load rate exceeds the maximum allowable transmission load rate. Among them, R OL,i For the line overload risk in the i-th time period; l j,k For system state E k Load rate of line j; l j,max N represents the maximum load rate allowed for transmission on branch j. L This represents the total number of branch roads. Power backfeed risk: calculated based on the percentage of reverse power at the starting line relative to the line's operating limit. Among them, R PR,i For the power backfeed risk in the i-th time period; l PR,k For system state E k The percentage of the power of the downstream line relative to the operating limit of the power grid transformer or line; if reverse transmission occurs, the power is a negative value. The method for calculating the risk index of emergency scenario Nk at each time segment is as follows: Node voltage deviation risk: Using the baseline voltage before the fault as a reference, calculate the deviation of the node voltage after the fault and map it to severity. Wherein, the voltage at the node after disconnection is V. j,k V0 represents the node voltage value before the system disturbance, R j,k It is determined by the degree to which the node voltage exceeds the limit before the disturbance. Branch load change risk: Calculate the degree of change in branch load rate after a fault relative to the load rate before the fault. Among them, the load rate after disconnection is l j,k The branch load factor before the system disturbance is l0, R j,k It is determined by the degree of overload of the branch before the disturbance. N-1 pass rate: This refers to the percentage of system states that meet the requirements for node voltage and line load rate across all generated emergency combinations. It mainly includes the line N-1 pass rate and the node N-1 pass rate. Among them, R LC,i Let l be the route N-1 completion rate in the i-th time period; LC,k For state E k The following line ratio satisfies N-1, B LC,k For state E k The following satisfies the ratio of N-1 nodes.

5. The multi-scenario risk assessment method for a novel power distribution system according to claim 1, wherein step 4) involves constructing an objective weighting model based on the entropy weight method, including: Construct the probability distribution matrix of risk indicators: Calculate the probability that the j-th indicator falls within the level range of the i-th indicator. Where, x ij This represents the probability that the j-th indicator falls within the level range of the i-th indicator. The information entropy of each indicator is calculated based on the relative frequency of the indicator data. The smaller the information entropy, the greater the variability of the indicator. Objective weights are calculated based on information entropy, and the objective weights of indicators are inversely proportional to their information entropy: Among them, H j Let be the entropy weight of the j-th index; 1-E j This represents the degree of variability corresponding to the j-th indicator. The larger the information entropy, the smaller its degree of variability, and the smaller its impact on the overall risk assessment of the power distribution system. The construction steps of the subjective weighting model based on the analytic hierarchy process are as follows: First, a decision matrix P is established, using a scale of 1 to 9 to measure the relative importance between any two risk indicators. Then, the largest eigenvalue of the decision matrix and its corresponding eigenvector are calculated. Finally, matrix P is normalized to obtain the relative weights W of the decision indicators using the analytic hierarchy process (AHP). j . Where, λ max represents the largest eigenvalue of the decision matrix; n is the order of the decision matrix. When the consistency ratio CR is less than 0.1, it indicates that the overall consistency of the hierarchical ranking results under subjective thinking is relatively high. RI is the random consistency index for the corresponding n-dimensional matrix, and RI increases with the increase of the matrix order. A consistency check is performed on the decision matrix; when the consistency ratio is less than 0.1, the eigenvectors are normalized and used as subjective weights. After obtaining the weights of the subjective and objective indicators, the combined weights of each indicator are obtained using the following method: Where n is the number of indicators; H j W j These represent the subjective and objective weights of the j-th indicator, respectively. The combined weight coefficients are calculated, then normalized. Finally, the final weight value is obtained by linearly calculating the coefficients with the subjective and objective weights.

6. The multi-scenario risk situation assessment method for a new type of power distribution system according to claim 1, wherein the specific process of step 5) is as follows: multiply the risk impact coefficient, the normalized probability index and the combined weight, and obtain the power distribution network risk situation assessment value E under complex risk scenarios by weighted summation. in, a ij x represents the impact coefficient of different risk levels on the power distribution system. ij S is a normalized probability index. j The weights are then combined. The risk status of the power distribution system is then classified into three levels: safe, warning, and dangerous.

Images