A Grid Risk Assessment Method Applicable to Photovoltaic Grid-Connected Access Systems

Abstract

The present invention belongs to the technical field of power system risk assessment, and discloses a power grid risk assessment method applicable to a photovoltaic grid-connected access system. By proposing photovoltaic fluctuation bearing capacity index PVCI, photovoltaic-load mismatch risk index SLMRI, photovoltaic reactive power support capacity index PVQSI, and node voltage quality assessment index NYQI, this method centrally reflects the impacts of photovoltaic grid-connected power generation on the overall operation status of the power grid, the mismatch with the load, the voltage regulation effect, and the voltage stability of system nodes. Through the multi-dimensional game entropy weight combination weighting method, the information entropy weight, multi-dimensional boundary contribution degree, and game contribution degree among various indicators are calculated to obtain the comprehensive weight of each indicator. Combining with the collected indicator data, the risk value of the power grid is finally calculated. The present invention measures the importance of indicators from three perspectives of information volume, boundary contribution, and synergy, combines objective weighting and subjective weighting, and improves the accuracy of power grid risk assessment under photovoltaic grid-connected conditions.

Description

Technical Field

[0001] The invention belongs to the technical field of power system risk assessment, and specifically relates to a power grid risk assessment method applicable to a photovoltaic grid-connected access system. Background Art

[0002] With the green transformation and structural adjustment of the power energy industry, more and more photovoltaic power generation is connected to the grid, and the proportion of new energy power generation in the power grid is increasing day by day. Affected by natural weather conditions and other factors, photovoltaic power generation has characteristics such as randomness and volatility. Thus, the volatility problems brought about by a large number of photovoltaic grid connections have a significant impact on the safe and stable operation of the power system, and its proportion in power production is getting larger and larger. Traditional power grid risk assessment methods do not fully consider the impact brought about by the random volatility of photovoltaic power output, and there are still deficiencies in the rationality and comprehensiveness of the risk assessment of new power systems under high-proportion new energy grid connections. Summary of the Invention

[0003] The purpose of the invention is to provide a power grid risk assessment method applicable to a photovoltaic grid-connected access system, so as to realize the risk assessment of the power grid operation state while considering the fluctuating output of photovoltaic power, thereby providing a warning of the power grid operation state for power grid operators and a basis for judging dispatching.

[0004] To achieve the above purpose, the invention provides the following technical solutions:

[0005] A power grid risk assessment method applicable to a photovoltaic grid-connected access system includes the following steps:

[0006] Step 1: Calculate the output value of the photovoltaic power station according to the light intensity, set the data time resolution, calculate the power flow result of the power grid under the photovoltaic power output of each moment section, and save the output value of the photovoltaic power station and the power flow result of the power grid under each moment section as a sample in the database;

[0007] Step 2: Calculate the photovoltaic volatility bearing capacity index PVCI, the photovoltaic-load mismatch risk index SLMRI, the photovoltaic reactive power support capacity index PVQSI, and the node voltage quality assessment index NYQI, and perform positive normalization on the calculated original index values and save them in the sample sequence of the database under the corresponding moment section;

[0008] Step 3: Perform standardization processing on the positively normalized index data in Step 2, then calculate the information entropy weight, the game weight, and the boundary contribution weight, and synthesize the three weights to obtain the comprehensive weight of each index;

[0009] Multiply the normalized comprehensive weight by the standardized index data, and use the weighted summation method to obtain the unnormalized power grid operation risk state value ; and normalize to obtain the final power grid operation risk status value , and based on judge the operation risk status of the overall power grid.

[0010] Compared with the prior art, the present invention has the following beneficial effects:

[0011] As described above, the present invention relates to a power grid risk assessment method applicable to a photovoltaic grid-connected access system. Through the photovoltaic fluctuation bearing capacity index PVCI, the photovoltaic-load mismatch risk index SLMRI, the photovoltaic reactive power support capacity index PVQSI, and the node voltage quality assessment index NYQI, it centrally reflects the impact of photovoltaic grid-connected power generation on the overall operation status of the power grid, the mismatch with the load, the voltage regulation effect, and the voltage stability of system nodes, and obtains the power grid operation risk status value through the multi-dimensional game entropy weight combination weighting method; the present invention measures the importance of indicators from three perspectives of information volume, boundary contribution, and synergy, combines objective weighting and subjective weighting, avoids the one-sidedness of a single method, and considers the variation law of samples in multi-dimensional space, improving the accuracy of power grid risk assessment under photovoltaic grid-connected conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for use in the embodiments.

[0013] Figure 1 is a flowchart of a power grid risk assessment method applicable to a photovoltaic grid-connected access system in an embodiment;

[0014] Figure 2 is a topological structure diagram of the IEEE39 node system in an embodiment;

[0015] Figure 3 is a change curve graph of the solar light intensity in 24 hours in an embodiment;

[0016] Figure 4 is a change situation graph of four indicators in 24 hours in an embodiment;

[0017] Figure 5 is a bar distribution graph of the weight values of four indicators in an embodiment;

[0018] Figure 6 is a change curve graph of the power grid operation risk status value in 24 hours in an embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments.

[0020] All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0021] Embodiment

[0022] As Figure 1 shown, this embodiment describes a power grid risk assessment method applicable to a photovoltaic grid-connected access system. This method is aimed at the scenario of fluctuating power output of photovoltaic grid-connected power generation in the power system. By proposing evaluation indicators considering the photovoltaic fluctuating power output and its impact on the power grid, the multi-dimensional game entropy weight combination weighting method is used to calculate the weights of the indicators, and combined with the indicator data, the weighted power grid operation risk status value is obtained. Based on this, the operation safety status of the power grid can be judged, so as to provide early warnings of the system operation status for power grid operation personnel and provide a basis for judging dispatching.

[0023] This method includes the following steps:

[0024] Step 1: Calculate the power output value of the photovoltaic power station according to the light intensity, set the data time resolution, calculate the power flow result of the power grid under the photovoltaic power generation output at each moment, and save the power output value of the photovoltaic power station and the power flow result at each moment section as a sample in the database.

[0025] For the large-scale grid connection of photovoltaic power stations, the power output of the entire station is:

[0026] ;

[0027] In the formula, is the simulated light intensity in the area where the photovoltaic power station is located, is the total array area of the photovoltaic panels of the entire photovoltaic power station, is the conversion efficiency of the photovoltaic inverter, is the conversion efficiency of the photovoltaic panel, and its mathematical expression is:

[0028] ;

[0029] In the formula, represents the light intensity threshold.

[0030] In this embodiment, the data time resolution is set to 30 s, that is, the power generation output of the entire network is recorded every 30 s, the power flow operation result of the power grid under this power output is calculated, and the photovoltaic power generation output value and the power flow result at the moment section are saved as a sample in the database.

[0031] Step 2: According to the data saved in the database in Step 1, calculate the Photovoltaic Fluctuation Carrying Capacity Index (PVCI), the Photovoltaic-Load Mismatch Risk Index (SLMRI), the Photovoltaic Reactive Power Support Capacity Index (PVQSI), and the Node Voltage Quality Assessment Index (NYQI). After normalizing the calculated original index values, save them in the sample sequence of the database under the corresponding time section.

[0032] This embodiment proposes the Photovoltaic Fluctuation Carrying Capacity Index (PVCI), the Photovoltaic-Load Mismatch Risk Index (SLMRI), the Photovoltaic Reactive Power Support Capacity Index (PVQSI), and the Node Voltage Quality Assessment Index (NYQI) for calculating the grid operation risk assessment under the condition of photovoltaic output fluctuation.

[0033] (1) Photovoltaic Fluctuation Carrying Capacity Index (PVCI)

[0034] This index is used to measure the grid's ability to cope with the fluctuating output of photovoltaic power. It evaluates the stability and flexibility of the grid in coping with the fluctuating output of photovoltaic power by comparing the change rate of photovoltaic output with the adjustment ability of the grid's adjustable resources.

[0035] The calculation formula of the Photovoltaic Fluctuation Carrying Capacity Index (PVCI) is:

[0036] ;

[0037] In the formula, , which represents the difference in the output values of the photovoltaic power station at time and at time. The time interval is taken as 30 s (the time interval depends on the data time resolution. The data time resolution in this embodiment is 30 s, that is, the photovoltaic output and the grid power flow operation data are collected every 30 s); represents the maximum available capacity of the adjustable resources in the grid at time, including adjustable generating units, etc.; represents the minimum technical output of the adjustable resources at time; represents the adjustment ability of the adjustable resources in the grid, that is, the maximum adjustment range that the whole network's power generation resources can provide.

[0038] (2) Photovoltaic-Load Mismatch Risk Index (SLMRI)

[0039] This index is used to quantify the mismatch risk between the photovoltaic power generation output and the power load. It represents the risk level of the grid in coping with this mismatch by calculating the dynamic difference between the photovoltaic output and the load.

[0040] The calculation formula of the photovoltaic-load mismatch risk index SLMRI is as follows:

[0041] ;

[0042] In the formula, and respectively represent the load values of the power grid at the time of and ; and respectively represent the output values of the photovoltaic power station at the time of and ; is the time period weight factor, which is used to reflect the difficulty of power grid scheduling in different time periods (the time period weight factor can be set according to the specific situation of the power grid. For example, the weight in the peak electricity consumption period can be higher); is the evaluation period, which is set to the data time resolution in this embodiment.

[0043] (3) Photovoltaic reactive power support ability index PVQSI

[0044] This index is used to measure the ability of the photovoltaic power generation system to provide reactive power support in the power grid. It reflects the reactive power support ability and level of the photovoltaic power station connected to the system by calculating the weighted value of the reactive power provided by the photovoltaic power station and the node voltage deviation rate of the corresponding power station.

[0045] The calculation formula of the photovoltaic reactive power support ability index PVQSI is as follows:

[0046] ;

[0047] In the formula, is the reactive power provided by the photovoltaic power station at node ; and are the upper and lower limits of reactive power regulation of the photovoltaic inverter at the current active power level; is the absolute value of the voltage deviation of the node where the photovoltaic power station is connected; is the maximum allowable voltage deviation limit of node , generally taken as 0.05.

[0048] (4) Node voltage quality assessment index NYQI

[0049] This index is used to measure the comprehensive index of the voltage quality of each node in the power system. It reflects the stability and overall quality of the system voltage by calculating the deviation of the node voltage from the rated voltage and weighting it with the importance of the node.

[0050] The calculation formula of the node voltage quality assessment index NYQI is as follows:

[0051] ;

[0052] Wherein, represents the actual operating voltage of node ; represents the rated operating voltage of node ; represents the weight coefficient of node , which is used to reflect the importance of the node. The weight coefficient can be determined according to factors such as the load importance or power generation capacity of the node; represents the number of nodes in the system.

[0053] For these four indicators, positive and negative indicator types are defined. Among them, the photovoltaic fluctuation carrying capacity index PVCI, the photovoltaic-load mismatch risk index SLMRI, and the node voltage quality assessment index NYQI are positive indicators, that is, the larger the indicator value, the higher the risk operation state of the power grid; the photovoltaic reactive power support ability index PVQSI is a negative indicator, that is, the smaller the indicator value, the higher the risk operation state of the power grid.

[0054] For the negative indicator , positive transformation is carried out to facilitate the later weighted processing of indicators. The calculation formula for positive transformation is:

[0055] ;

[0056] Wherein, is the original indicator value of the negative indicator, is the indicator value after positive transformation.

[0057] For each calculated original indicator value, after necessary positive transformation, it is saved in the sample sequence of the database under the corresponding time section.

[0058] Step 3: Use the multi-dimensional game entropy weight combination weighting method to comprehensively weight the four indicators in Step 2, and add weights to each indicator value to obtain the power grid operation risk state value, from which the overall operation risk state of the power grid can be judged.

[0059] First, standardize the data of each indicator, then calculate the information entropy weight, game weight, and boundary contribution weight, and synthesize the three weights to obtain the comprehensive weight; then, multiply the comprehensive weight after normalization by the standardized data of each indicator (sample indicator value) and perform weighted summation to obtain the unnormalized power grid operation risk state value , and after normalization, the final power grid operation risk state value can be obtained , according to It can judge the overall operating risk state of the power grid.

[0060] Step 3.1. Data standardization

[0061] To eliminate the influence of the dimension of each index data and make different indexes comparable, map the data of each index to the interval, and the calculation formula is as follows:

[0062] ;

[0063] In the formula, is the original value of the th sample in the database on the th index; is the value after standardization; is the minimum value of the th index; is the maximum value of the th index; Through standardization processing, map the data of each index to the interval.

[0064] Step 3.2. Calculate the information entropy weight

[0065] (1) Calculate the probability matrix. Convert the standardized index values into probability forms to prepare for subsequent information entropy calculation. The probability values represent the relative importance of the samples on this index. The elements of the probability matrix are calculated as follows:

[0066] ;

[0067] In the formula, is the probability value of the th sample on the th index, is the sum of the standardized values of the th index, is the total number of samples.

[0068] (2) Calculate the information entropy. The greater the information entropy, the higher the uncertainty of the index and the less effective information it provides. The information entropy calculation formula is as follows:

[0069] ;

[0070] In the formula, is the information entropy of the th index; is the natural logarithm of the number of samples, which is a normalization factor; is the basic calculation unit of the amount of information.

[0071] (3)Calculate the information utility degree. The information utility degree represents the certainty degree of an indicator, that is, the ability to provide effective information. The higher the utility degree, the stronger the ability of the indicator to distinguish samples. The calculation formula of the information utility degree is as follows:

[0072] ;

[0073] In the formula, is the information utility degree of the th indicator, is the information entropy of the th indicator.

[0074] (4)Calculate the information entropy weight. The higher the information utility degree of an indicator, the higher its weight, indicating that the indicator has a stronger ability to distinguish the evaluation object. The calculation formula of the information entropy weight is as follows:

[0075] ;

[0076] In the formula, is the information entropy weight of the th indicator, is the sum of the information utility degrees of all indicators, is the total number of indicators.

[0077] Step 3.3. Calculate the game weight

[0078] (1)The basic formula of the Shapley value (game contribution degree). Based on the cooperative game theory, the Shapley value calculates the average marginal contribution of each indicator to the overall evaluation, considering the synergy between indicators. The basic formula is as follows:

[0079] ;

[0080] In the formula, is the Shapley value of the th indicator, is the subset of indicators that does not include the indicator , is the set of all indicators, is the number of indicators in the subset , is the total number of indicators, is the coalition value function of the indicator subset , is the marginal contribution of the indicator to the coalition (that is, the indicator subset ).

[0081] (2)Calculate the coalition value function. The coalition value function comprehensively considers the variation degree and correlation of indicators. The coefficient of variation Reflects the dispersion degree of the indicators, This item considers the redundancy between indicators. The lower the correlation, the larger this value, indicating a higher independence between indicators and more non-redundant information provided.

[0082] The calculation formula of the coalition value function is:

[0083] ;

[0084] In the formula, is the coalition value of the indicator subset ; is the coefficient of variation of the th indicator, where , is the standard deviation of the th indicator, is the average value of the th indicator; is the correlation coefficient matrix between the indicators in the subset ; is the identity matrix; is the average value of the absolute value of the difference between the correlation coefficient matrix and the identity matrix.

[0085] (3)Calculate the game weights. Normalize the Shapley values of each indicator to obtain the weights. The higher the Shapley value of an indicator, the higher its game weight, indicating a greater contribution of this indicator to the overall evaluation.

[0086] The calculation formula of the game weight is:

[0087] ;

[0088] In the formula, is the game weight of the th indicator, is the Shapley value of the th indicator, is the sum of the Shapley values of all indicators.

[0089] Step 3.4. Calculate the boundary contribution weights

[0090] (1)Calculate the gradient. Use the central difference method to approximately calculate the gradient of the sample in the direction of each indicator, indicating the rate of change of the sample in this direction. When the sample is a boundary point (such as or ), use the forward or backward difference method. The gradient calculation formula is as follows:

[0091] ;

[0092] In the formula, is the gradient of the rd sample on the th index, is the rd sample's normalized value on the th index, is the rd sample's normalized value on the th index.

[0093] (2) Calculate the sum of squared gradients. The sum of squared gradients reflects the degree of change of an index in the sample space. The larger the sum of squared gradients, the greater the impact of the index on the sample distribution and the more significant its contribution to boundary recognition. The formula for calculating the sum of squared gradients is as follows:

[0094] ;

[0095] In the formula, is the sum of squared gradients of the th index, is the squared gradient of the rd sample on the th index, is the total number of samples.

[0096] (3) Calculate the boundary contribution weight. Normalize the sum of squared gradients of each index to obtain the weight. The larger the sum of squared gradients of an index, the higher its boundary contribution weight, indicating that the index makes a greater contribution to the sample boundary division. The formula for calculating the boundary contribution weight is as follows:

[0097] ;

[0098] In the formula, is the boundary contribution weight of the th index, is the sum of squared gradients of the th index, is the sum of the sum of squared gradients of all indices, is the total number of indices.

[0099] Step 3.5, Calculate the comprehensive weight

[0100] (1) According to the information entropy weight, game weight, and boundary contribution weight that have been calculated, use the weighted summation method to combine the three weights to obtain the comprehensive weight. The formula for calculating the comprehensive weight of the th index is as follows:

[0101] ;

[0102] In the formula, is the The comprehensive weight of an index without normalization, is the information entropy weight of the th index, is the game weight of the th index, is the boundary contribution weight of the th index, is the proportionality coefficient of the information entropy weight, is the proportionality coefficient of the game weight, is the proportionality coefficient of the boundary contribution weight;

[0103] Among them, , and need to satisfy the constraint conditions:

[0104] ;

[0105] ;

[0106] (2) To ensure that the final weight satisfies the constraint condition that the sum of weights is 1, it is necessary to perform normalization processing on

[0107] ;

[0108] In the formula, is the comprehensive weight of the th index after normalization, is the comprehensive weight of the th index without normalization.

[0109] Step 3.6. Calculate the power grid operation risk status value

[0110] (1) Multiply the comprehensive weight after normalization by the standardized data (sample index values) of each index, and perform weighted summation to obtain the unnormalized power grid operation risk status value , and its calculation formula is:

[0111] ;

[0112] In the formula, is the standardized sample index value of the th sample in the database on the th index.

[0113] (2) Perform normalization processing on , map the unnormalized power grid operation risk status value to the interval, and the final power grid operation risk status value can be obtained. Through The operating risk status of the power grid can be judged by its magnitude, The larger it is, that is, the closer the value of is to 1, the higher the risk level of the power grid, and the more dangerous the operating state of the system; conversely, the smaller it is, that is,

[0114] A power grid risk assessment method applicable to the photovoltaic grid-connected access system described in this embodiment focuses on the impact of the fluctuation characteristics of photovoltaic power generation output on the system operating risk status, providing an important basis for formulating prevention and control strategies and emergency plans.

[0115] This embodiment takes into account that in the photovoltaic grid-connected system, its grid-connected scale is continuously expanding and the power generation proportion is continuously increasing, and its power generation capacity is extremely affected by natural environmental conditions, resulting in the characteristics of randomness and volatility of the output, which will pose serious threats and challenges to the safe and stable operation of the power grid.

[0116] Therefore, by proposing the photovoltaic fluctuation bearing capacity index PVCI, the photovoltaic-load mismatch risk index SLMRI, the photovoltaic reactive power support capacity index PVQSI, and the node voltage quality assessment index NYQI, to centrally reflect the impact of photovoltaic grid-connected power generation on the overall operating condition of the power grid, the mismatch with the load, the voltage regulation effect, and the node voltage stability of the system; through the multi-dimensional game entropy weight combination weighting method, calculate data such as the information entropy weight, multi-dimensional boundary contribution degree, and game contribution degree between each index of each index, and finally obtain the comprehensive weight of each index. The weighting method of the present invention measures the importance of the index from three angles of information volume, boundary contribution, and synergy, combines objective weighting and subjective weighting, not only avoids the one-sidedness of a single method, but also considers the change law of the sample in the multi-dimensional space, making the method more suitable for dealing with the evaluation index system with complex correlation relationships between indexes.

[0117] Next, the IEEE 39-node system is used as an example to verify the above method to prove the effectiveness and rationality of the method of this embodiment.

[0118] Such as Figure 2As shown in the figure, it is the topological structure diagram of the IEEE 39-bus system. This system consists of 39 bus nodes at the 345 kV level, 10 generator sets, and 46 transmission lines. The total installed capacity of the whole network reaches 6297.871 MW, the total active load of the whole network is 6253.23 MW, and the reactive load is 1387.1 MVar. In this embodiment, the power generation plants at the 32nd and 35th bus nodes in the original system are replaced by photovoltaic power stations according to equal power generation capacity, and photovoltaic power stations are newly added at the 4th, 15th, 25th, and 30th bus nodes, with corresponding capacities of 200 MW, 210 MW, 190 MW, and 205 MW respectively. Flexible adjustment resources are set at the 8th, 18th, 30th, 32nd, and 35th bus nodes, and the load of the whole network is set to fluctuate according to the general load curve rule.

[0119] In this embodiment, a 24-hour random simulation is carried out on the IEEE 39-bus system. The 24-hour light intensity change is randomly simulated and generated, and the light intensity value is recorded every 30 s and saved to the database. According to the collected light intensity data, the conversion efficiency calculation formula of the photovoltaic panel described in this embodiment is used:

[0120] ;

[0121] Calculate the conversion efficiency of the photovoltaic panel under different light intensities.

[0122] Then, the power output calculation formula of the photovoltaic power station is used:

[0123] ;

[0124] Calculate the power output of the photovoltaic power station under different light intensities. As Figure 3 shown, it is the 24-hour solar light intensity change curve.

[0125] In this embodiment, the solar light intensity values corresponding to 22 moments are selected. Among them, samples a~k are the recorded values of the previous sampling moment of samples 1~11 respectively, and the corresponding power output values of the photovoltaic power stations of the whole network are calculated and saved to the database. The 22 sample values are shown in Table 1.

[0126] Table 1: Sample values of light intensity and power output of the photovoltaic power stations of the whole network at 22 moment sections

[0127]

[0128] According to the power output situation of the photovoltaic power station, the power grid power flow at the corresponding moment section is calculated and saved to the database.

[0129] According to the output of the PV power station recorded in the database and the power grid flow, the index values of the samples numbered 1 to 11 at the corresponding cross-section at the corresponding moment are calculated, as shown in Table 2. Figure 4 It is the change of each index value during a day.

[0130] Table 2: Each index value under 11 moment cross-sections

[0131]

[0132] The comprehensive weight value of each index is calculated by using the multi-dimensional game entropy weight combination weighting method. Among them, the information entropy weight, Shapley value, multi-dimensional boundary contribution degree and comprehensive weight value are shown in Table 3. Figure 5 It is the weight bar chart of each index.

[0133] Table 3: Comprehensive weight values of each index

[0134]

[0135] The negative type indicators are positive processed, and the statistical results of each index value are normalized, and then weighted and summed with each comprehensive weight to obtain the power grid operation risk state value at different moments. Table 4 shows the power grid operation risk state values calculated in this embodiment under 11 moment cross-sections. Figure 6 It is the change curve of the power grid operation risk state value in 24 hours of a day in this embodiment.

[0136] Table 4: Power grid operation risk state values corresponding to 11 sample moments

[0137]

[0138] As can be seen from Table 1 and Table 4, among samples 1 to 11, as the overall intensity of solar irradiance continuously increases, the proportion of photovoltaic power generation becomes larger and larger, and with the influence of simulated random factors such as cloud occlusion on the irradiance, slightly larger power fluctuations will cause overload of key nodes or key lines, resulting in an increasingly high level of power grid operation risk; from the index values corresponding to each sample in Table 2, the larger the values of the photovoltaic-load mismatch risk index SLMRI and the photovoltaic reactive power support ability index PVQSI, the higher the power grid operation risk state of the corresponding sample, which conforms to the definition of the index type. Although the values of the photovoltaic fluctuation bearing capacity index PVCI and the node voltage quality evaluation index NYQI have relatively strong fluctuations, they also conform to the definition of the index type in the comparison of index values at adjacent moment cross-sections; from Figure 6It can be seen that the variation law of the grid operation risk state value during a day basically fits the fluctuation variation law of the solar light intensity corresponding to the same day. For example, in the time period from 10:19:00 to 10:28:00, the solar light intensity fluctuates and suddenly drops from 744.78 W / m2 to 83.78 W / m2, and the corresponding PV output of the whole network suddenly drops from 1873.35 MW to 190 MW, resulting in a large amount of power deficit and drastic changes in the power flow distribution in the grid, causing a strong change in the grid risk state, corresponding to Figure 6 the highest peak in Figure 6 is the operation condition of this time period. Thus, it can be verified that the system risk state assessment carried out according to the method of the present invention is consistent with the actual operation state of the grid, and therefore the effectiveness and rationality of the present invention are verified.

[0139] It can also be intuitively obtained from the system states at 22 moment cross-sections that in a system with PV grid connection, the higher the proportion of PV output in the power generation, the greater the volatility of the output, the higher the risk level of the system operation state will be, the more dangerous the risk situation faced by the system will be, and the greater the possibility of accidents such as power outages; when the PV output is insufficient or drops suddenly in a short period of time, it will lead to serious imbalance in power transmission in the system and large-scale transfer of power flow, serious node voltage over-limit phenomenon, serious overload of some transmission lines, and sharp increase in the line load rate, resulting in jumper and causing the grid to trip and shut down; when the short-term fluctuation amplitude of the PV output change rate is large, the operation state of the system develops towards a high-risk level with the increase of the change rate. Also, it can be seen from the comprehensive weight values obtained in Table 3 that the weight of the PV fluctuation bearing capacity index PVCI is relatively high, and the node voltage quality assessment index NYQI is the second. This shows that PV grid connection has a strong impact on the overall operation state of the system and the stability of the node voltage. Therefore, the factor of PV fluctuating output cannot be ignored when conducting risk assessment on the grid.

[0140] The embodiments of the present invention are only used to illustrate the technical solutions of the present invention and not to limit them. For those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirits of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A grid risk assessment method applicable to a photovoltaic grid-connected access system, characterized in that, It includes the following steps: Step 1: Calculate the output value of the photovoltaic power station according to the light intensity, set the data time resolution, calculate the power flow result of the power grid under the photovoltaic power generation output at each moment, and save the output value of the photovoltaic power station and the power flow result of the power grid at each moment section as a sample in the database; Step 2: Calculate the photovoltaic volatility carrying capacity index PVCI, the photovoltaic-load mismatch risk index SLMRI, the photovoltaic reactive power support ability index PVQSI, and the node voltage quality assessment index NYQI. The calculation formulas are as follows: where, ΔP pv (t) = P pv (t) - P pv (t - Δt), which represents the difference in the output values of the PV power station at time t and time (t - Δt), and the time interval Δt depends on the data time resolution; P flexible (t) represents the maximum available capacity of adjustable resources in the power grid at time t, including adjustable generating units; P min_flexible (t) represents the minimum technical output of adjustable resources at time t; ∑(P flexible (t) - P min_flexible (t)) represents the adjustment capacity of adjustable resources in the power grid, that is, the maximum adjustment range that the power generation resources of the whole network can provide; where P load (t) and P load (t - Δt) represent the load values of the power grid at time t and t - Δt, respectively; P pv (t) and P pv (t - Δt) represent the output values of the PV power station at time t and t - Δt, respectively; ω(t) is the time period weight factor, which is used to reflect the difficulty of power grid scheduling in different time periods; T is the evaluation period; where Q i,PV,actual is the reactive power provided to the PV power station at node i; Q i,PV,max and Q i,PV,min are the upper and lower limits of reactive power regulation of the PV inverter at the current active power level; |ΔU i | is the absolute value of the voltage deviation at the node where the PV power station is connected; ΔU i,max is the maximum allowable voltage deviation limit at node i. where U i represents the actual operating voltage of node i; U ir represents the rated operating voltage of node i; ω i represents the weight coefficient of node i, which is used to reflect the importance of the node; n represents the number of nodes in the system; Perform positive normalization on the calculated original index values and save them in the sample sequence of the database at the corresponding moment section; Step 3: Perform standardization processing on the index data after positive normalization in Step 2, then calculate the information entropy weight, the game weight, and the boundary contribution weight, and synthesize the three weights to obtain the comprehensive weight of each index; Multiply the comprehensive weight after normalization by the index data after standardization, and use the weighted summation method to obtain the unnormalized power grid operation risk state value E'; perform normalization on E' to obtain the final power grid operation risk state value E, and judge the overall operation risk state of the power grid according to E.

2. The grid risk assessment method applicable to a photovoltaic grid-connected access system according to claim 1, characterized in that The calculation process of the output value of the photovoltaic power station in Step 1 is as follows: For the large-scale grid connection of the photovoltaic power station, the output of the entire station is: P solar = MAηη inv ; Where M is the simulated light intensity in the area where the PV power station is located, A is the total array area of the PV panels of the entire PV power station, η inv is the conversion efficiency of the PV inverter, and η is the conversion efficiency of the PV panels, and its mathematical expression is: Where M k represents the light intensity threshold value.

3. A grid risk assessment method applicable to a photovoltaic grid-connected access system according to claim 1, characterized in that, The process of positive normalization of each original index value in Step 2 is as follows: First, define the photovoltaic volatility carrying capacity index PVCI, the photovoltaic-load mismatch risk index SLMRI, and the node voltage quality assessment index NYQI as positive type indexes, and the photovoltaic reactive power support ability index PVQSI as a negative type index; Then, perform positive normalization on the negative type index. The formula is: x i ′ = max(x i ) - x i ; where x i is the original index value of the negative index, and x i ' is the index value after positive transformation.

4. A grid risk assessment method applicable to a photovoltaic grid-connected access system according to claim 1, characterized in that The formula for standardizing the index data after positive normalization in Step 3 is: where x ij is the original value of the i-th sample in the database on the j-th index; x' ij is the value after standardizing x ij ; min(x j ) is the minimum value of the j-th index; max(x j ) is the maximum value of the j-th index; through standardization processing, the data of each index is mapped to the interval [0, 1].

5. A grid risk assessment method applicable to a photovoltaic grid-connected access system according to claim 1, characterized in that, The calculation process of the information entropy weight in Step 3 is as follows: (1) Convert the standardized index value into a probability form. The calculation formula for each element of the probability matrix is: where p ij is the probability value of the i-th sample on the j-th index, is the sum of the standardized values of the j-th index, m is the total number of samples; x ij is the original value of the i-th sample in the database on the j-th index; x' ij is the value after standardizing x ij ; (2) Calculate the information entropy. The calculation formula is: where, e j is the information entropy of the j-th index; ln(m) is the natural logarithm of the number of samples, which is a normalization factor; p ij ln(p ij ) is the basic calculation unit of the amount of information; the larger the information entropy, the higher the uncertainty of the index and the less effective information it provides; (3) Calculate the information utility degree. The calculation formula is: d j = 1 - e j ; where d j is the information utility degree of the j-th index, and e j is the information entropy of the j-th index; the information utility degree represents the degree of certainty of the index, that is, the ability to provide effective information. The higher the utility degree, the stronger the ability of the index to distinguish samples; (4) Calculate the information entropy weight. The calculation formula is: In the formula, is the information entropy weight of the j-th index, is the sum of the information utility degrees of all indexes, and n is the total number of indexes; the higher the information utility degree of an index, the higher its weight, indicating that the index has a stronger ability to distinguish the evaluation object.

6. The grid risk assessment method applicable to a photovoltaic grid-connected access system according to claim 1, wherein The calculation process of the game weight in Step 3 is as follows: (1) Calculate the Shapley value. The basic formula is as follows: where φ j is the Shapley value of the j-th index, S is the subset of indices that does not include index j, N is the set of all indices, |S| is the number of indices in subset S, n is the total number of indices, v(S) is the coalition value function of index subset S, and v(S ∪ {j}) - v(S) is the marginal contribution of index j to index subset S; (2) Calculate the coalition value function. The calculation formula is: where \(v(S)\) is the coalition value of the index subset \(S\); \(CV\) j is the coefficient of variation of the \(j\)-th index, where \(\sigma\) j is the standard deviation of the \(j\)-th index, and \(\mu\) j is the mean value of the \(j\)-th index; \(R\) is the correlation coefficient matrix among the indices in the subset \(S\); \(I\) is the identity matrix; is the average value of the absolute value of the difference between the correlation coefficient matrix and the identity matrix; (3) Calculate the game weight. The calculation formula is: In the formula, is the game weight of the j-th index, and φ j is the Shapley value of the j-th index, is the sum of the Shapley values of all indices.

7. A grid risk assessment method applicable to a photovoltaic grid-connected access system according to claim 1, characterized in that The calculation process of the boundary contribution weight in Step 3 is as follows: (1) Use the central difference method to calculate the gradient of the sample in each index direction. When the sample is a boundary point, use the forward difference method or the backward difference method. The gradient calculation formula is as follows: In the formula, is the gradient of the i-th sample on the j-th index, and x′ i+1,j is the standardized value of the (i + 1)-th sample on the j-th index, and x′ i-1,j is the standardized value of the (i - 1)-th sample on the j-th index; (2) Calculate the sum of the squares of the gradients. The calculation formula is as follows: where g j is the sum of squared gradients of the j-th index, is the squared gradient of the i-th sample on the j-th index; (3) Calculate the boundary contribution weight. The calculation formula is as follows: In the formula, is the boundary contribution weight of the j-th index, and g j is the sum of the squares of the gradients of the j-th index, is the sum of the sums of the squares of the gradients of all indices.

8. A grid risk assessment method applicable to a photovoltaic grid-connected access system according to claim 1, characterized in that The process of calculating the comprehensive weight and normalizing it in Step 3 is as follows: (1) Calculate the comprehensive weight by using the weighted summation method. The calculation formula is: where ω j is the unnormalized comprehensive weight of the j-th index, is the information entropy weight of the j-th index, is the game weight of the j-th index, is the boundary contribution weight of the j-th index, α is the proportionality coefficient of the information entropy weight, β is the proportionality coefficient of the game weight, and γ is the proportionality coefficient of the boundary contribution weight; Among them, α, β, and γ need to meet the constraint conditions: α + β + γ = 1; α, β, γ ≥ 0; (2) Perform normalization on the comprehensive weight. The calculation formula is: In the formula, is the comprehensive weight of the j-th index after normalization, and ω j is the comprehensive weight of the j-th index without normalization.

9. A grid risk assessment method applicable to a photovoltaic grid-connected access system according to claim 1, characterized in that, The calculation process of the final power grid operation risk status value E in step 3 is as follows: (1) Calculate the unnormalized power grid operation risk status value E′, and its calculation formula is: where x′ ij is the standardized sample index value of the i-th sample on the j-th index in the database; (2) Perform normalization processing on E′, map E′ to the interval [0, 1] to obtain E; the closer the value of E is to 1, the higher the risk level of the power grid, and the more dangerous the operating state of the system; on the contrary, the closer the value of E is to 0, the lower the risk level of the power grid, and the safer the operating state of the system.

Images