A load regulation method based on high-voltage customer electricity external risk assessment

Abstract

This invention discloses a load control method based on external risk assessment of high-voltage customer electricity consumption, belonging to the field of load control technology. The method includes step A: acquiring a dataset of external risk indicators for high-voltage customers, as well as the subjective, objective, and independence weights of the indicator system, and constructing an indicator classification set based on the independence weight distribution; step B: establishing the constraints and objective function of a self-learning comprehensive weight model based on the independence weight criterion; step C: determining the risk assessment matrix and establishing an assessment and grading system combining the self-learning comprehensive weight model and the cloud model; and step D: formulating a load control strategy based on the external risk assessment of high-voltage customer electricity consumption. This invention improves the flexibility and reliability of the evaluation, ensuring the power safety of high-voltage customers during the controlled process and the balance of the power distribution network's source side.

Description

Technical Field

[0001] This invention relates to the field of load control technology, and in particular to a load regulation method based on external risk assessment of high-voltage customer electricity consumption. Background Technology

[0002] With the transformation and upgrading of traditional power systems to new power systems, significant changes have occurred in power supply structure and load patterns, resulting in a marked increase in the randomness and volatility of both the "source" and "load" sides. The proportion of new energy power generation, such as wind and solar power, has increased significantly across regions, and its output is closely coupled with weather changes, exhibiting considerable volatility. The uncertainty risks of traditional "source follows load" operation have increased, making it more prone to problems such as voltage and power exceeding limits, leading to increased difficulty in grid dispatching and operation. Maintaining source-load balance by flexibly allocating and controlling load resources such as high-voltage customers can effectively reduce these risks.

[0003] Currently, many studies employ combined weighting methods that integrate subjective and objective weighting to mitigate the evaluator's subjective bias. For example, when using entropy methods for objective weighting, the entropy method merely determines the weight of an indicator based on its dispersion, without considering the interrelationships between indicators, potentially leading to misjudgments. Therefore, some scholars have gradually introduced the independence weighting method into combined weighting methods. This method reflects the redundancy of information across indicators, further reducing information redundancy and making the combined weighting results more reasonable.

[0004] However, when controlling high-voltage customers, the actual situation of the power grid they are connected to should be considered, the impact of the power grid on high-voltage customers should be assessed, and potential risks should be identified to ensure that high-voltage customers can safely and reliably participate in the precise load regulation of the new power system. Therefore, a load regulation method based on the external risk assessment of high-voltage customers is needed to reasonably and effectively formulate load regulation decisions based on the risk assessment results. Summary of the Invention

[0005] The purpose of this invention is to propose a load control method based on external risk assessment of high-voltage customer electricity consumption, characterized in that the method includes the following steps:

[0006] Step A: Obtain the dataset of external risk indicators for high-voltage customers' electricity consumption, as well as the subjective, objective, and independence weights of the indicator system, and construct an indicator classification set based on the distribution of independence weights;

[0007] Step B: Establish the constraints and objective function of the self-learning integrated weight model based on the independence weight criterion;

[0008] Step C: Determine the risk assessment matrix and establish an assessment and rating system that combines a self-learning comprehensive weight model with a cloud model;

[0009] Step D: Develop a load control strategy based on an assessment of external risks associated with high-voltage customers' electricity consumption.

[0010] Step A specifically includes the following sub-steps:

[0011] Step A1: Use the fast sorting method for system states considering multi-state component models to obtain m different preset scenarios of the distribution network and their probability values; through power flow calculation, obtain the electrical dataset X of n index combinations under the m different preset scenarios. n|m ={X1,X2,…,X n};

[0012] Step A2: Using the analytic hierarchy process (AHP) and interval estimation optimization method, construct the discriminant matrix B based on expert experience. n Combined with electrical dataset X n|m ={X1,X2,…,X n Obtain the subjective weight values ​​of each indicator. Estimated range of subjective weights for each indicator The subjective weight matrix of the external risk index system for high-voltage customer electricity consumption is then obtained by combining these components.

[0013] Step A3: Using the entropy method, based on the electrical dataset X n|m ={X1,X2,…,X n Obtain the objective weight values ​​of each indicator. And combine them to obtain the objective weight matrix of the external risk index system for high-voltage customer electricity consumption.

[0014] Step A4: Using the independence method, based on the electrical dataset X n|m ={X1,X2,…,X n Obtain the independence weight values ​​for each indicator. The combination yields an independent objective weight matrix for the external risk indicators system of high-voltage customer electricity consumption.

[0015] Step A5, W du The weight value of each selected indicator is compared with the weight values ​​of the remaining unselected indicators, and the indicators are classified based on the objective weights of independence, resulting in a set of category labels H = {h1, h2, ..., hn} for n indicators. n}

[0016] In step A2, the estimation range of the subjective weights of each indicator is obtained. The formula is as follows:

[0017]

[0018] In the formula: Let B be the subjective weight value of the i-th indicator. n The discriminant matrix of the analytic hierarchy process (AHP) is... For a subjective weight matrix of n indicator systems;

[0019] Obtain the subjective weight values ​​of each indicator. The formula is as follows:

[0020]

[0021] Where E is the maximum entropy value corresponding to the maximum entropy criterion.

[0022] In step A3, the objective weight values ​​of each indicator are obtained. The formula is as follows:

[0023]

[0024] In the formula: m is the total number of preset scenes, P ij E represents the proportion of the i-th preset scenario under the j-th indicator. j Let the entropy value be the j-th index. Let P be the objective weight value of the j-th indicator, and set P as... ij When P = 0, ij lnP ij =0.

[0025] In step A4, the independence weight values ​​of each indicator are obtained. The formula is as follows:

[0026]

[0027] In the formula: R j Let be the multiple correlation coefficient of the j-th indicator. For X m|n The average value, For X m|n Eliminate the remaining matrix. Let be the objective weight of independence for the j-th indicator.

[0028] In step A5, a category label set H = {h1, h2, ..., h} of n indicators is obtained. n The formula for} is as follows:

[0029]

[0030]

[0031] In the formula: add is the i-th index x i weight value Greater than the k-th preset scenario indicator set X n|kThe number of times the weight values ​​of the remaining indicators are jian is the number of times the weight values ​​of the i-th indicator x are expressed. i weight value Less than the k-th preset scenario indicator set X n|k The number of times the weight values ​​of the remaining indicators are denoted by buq, where buq is the number of times the weight values ​​of the i-th indicator x are denoted by buq. i weight value Equal to the k-th preset scenario indicator set X n|k The number of times the remaining indicator weight values ​​are displayed; {1,2,3} correspond to the category labels of increasing, decreasing and uncertain types, respectively.

[0032] The constraints in step B include:

[0033] Subjective interval constraints:

[0034]

[0035] Indicator label type constraints:

[0036]

[0037] In the formula: These are the upper and lower limits of the weight range for indicator j, respectively.

[0038] The objective function in step B is:

[0039]

[0040] In the formula: Let β be the self-learning combined weight of the j-th indicator. j1 β is the combination coefficient of the subjective weights of the j-th indicator. j2 Let be the combination coefficient of the objective weights of the j-th indicator. The relative unit slope represents the independence weights among the indicators.

[0041] Step C specifically includes the following steps:

[0042] Step C1: Substitute the non-negatively normalized indicator dataset into equation (10) to calculate the membership matrix R of each indicator belonging to different risk levels.

[0043]

[0044] Step C2: Integrate the membership matrices R of each indicator to construct the risk assessment matrix R. ∑ :

[0045]

[0046] In the formula: n is the total number of preset scenes, r ij Let be the membership value of the i-th indicator belonging to risk level j;

[0047] Step C3: Define the self-learning combined weights based on the independence objective weight criterion as the indicator layer weight vector W. R The risk assessment matrix R determined by the cloud model ∑ Calculate and obtain the criterion-level risk assessment matrix G, which consists of the risk assessment results of each of the m preset scenarios. ∑ And use it as the risk assessment matrix at the criterion level;

[0048] Step C4: Divide the different preset scenario probability sets p(C i The summation and normalization are performed to obtain the criterion layer weight vector W. p ={p * (C1),p * (C2),…,p * (C m Combined with the risk assessment matrix G at the criterion level ∑ The target layer evaluation result Z = {z1, z2, z3, z4} is calculated.

[0049] Step C5: Based on the principle of maximum membership, select the maximum value z in Z. max The corresponding risk level is used to assess and classify the overall external risks of high-voltage customers' electricity consumption.

[0050] Step D specifically includes the following steps:

[0051] Step D1: Obtain a dataset X of power values ​​from both the source and load sides under 100 different preset scenarios. 100|2 ={X1,X2,…,X 100}, where X j ={P f ,P g}, P f P is the total power required by the distribution network load. g The total power that the generators in the distribution network can provide;

[0052] Step D2: Based on equation (12), filter the power distribution network source-load imbalance scenarios and form a scenario label set Q = {a1, a2, ..., a 100}:

[0053]

[0054] Step D3: Filter the scenes with a value of 1 from the scene label set Q to form the source-load imbalance scene set Q. S ={a1,a2,…,a s The self-learning risk assessment method in steps A to C is used to conduct external risk assessments on the safe electricity use of all high-voltage customers connected to the distribution network, forming a risk set U = {b1, b2, ..., b}.s}, where b1 represents the external risk level of safe electricity use for u high-voltage customers in scenario 1 of source-load imbalance, and b1 = {b 11 ,b 12 ,…,b 1u};

[0055] Step D4: Divide the adjustable high-voltage customer set based on the risk set U; set different adjustable risk level thresholds θ according to actual control needs:

[0056] Step D5: Compare the risk set U obtained in step D1 with the adjustable risk level threshold θ set in step D4 to obtain the adjustable high-voltage customer set K = {k1,k2,…,k s};

[0057] Step D6: Construct a load reduction strategy model based on the adjustable high-voltage customer set K;

[0058] Step D7: Solve for the total load that each adjustable high-voltage customer needs to control based on the load reduction strategy model in step D6.

[0059] The load reduction strategy model includes:

[0060] Objective function;

[0061]

[0062] Where C t This refers to the load reduction amount;

[0063] Constraints:

[0064]

[0065] Among them, P t (V,δ)=V t ∑V r (G tr cosδ tr +B tr sinδ tr );Q t (V,δ)=V t ∑V r (G tr cosδ tr -B tr sinδ tr );G th and B th δ represents the real and imaginary parts of the admittance matrix in row t and column h; V is the magnitude of the node bus voltage; δ is the phase angle difference between the two ends of the line; V t It is the voltage magnitude at node t; V rIt is the voltage magnitude at node r; δ tr It is the phase angle difference between nodes t and r; C t P is the load reduction at node t; LDt Q represents the active load at node t; LDt The reactive load at node t; and These are the upper and lower limits of injected active and injected reactive power at generator node t, respectively; TR h It is the actual transmission capacity of the transmission line h; Yan is the maximum transmission capacity of the transmission line h; V t max and V t min These are the upper and lower limits of the voltage amplitude of the bus at node t, respectively; G, NG, N, L, S and U are the sets of load bus nodes, generator bus nodes, all bus nodes, all transmission lines, source-load imbalance scenarios and adjustable high-voltage customers in the power transmission system, respectively.

[0066] The beneficial effects of this invention are as follows:

[0067] This invention ensures the advantages of the comprehensive evaluation method in integrating subjective and objective factors, and is more convenient and faster in handling constraints compared to traditional weighting methods. At the same time, it improves the flexibility and credibility of the evaluation, and ensures the power safety of high-voltage customers in the controlled process and the balance of the power distribution network source side. Attached Figure Description

[0068] Figure 1 This is a flowchart of the load control method based on external risk assessment of high-voltage customer electricity consumption according to the present invention;

[0069] Figure 2 This is a diagram of the IEEE-24-RTS distribution network topology.

[0070] Figure 3 This is a comparison chart of the combined weights of various indicators between the subjective and objective weighted evaluation method based on the product method and the method proposed in this invention.

[0071] Figure 4 This is a comparison chart of the relative unit slopes of various indicators of the subjective and objective weighted combination evaluation method based on the product method and the method proposed in this invention.

[0072] Figure 5 A comparison chart of the membership results of each risk level of high-voltage customer electricity consumption between the subjective and objective weighted combination assessment method based on the product method and the method proposed in this invention.

[0073] Figure 6 This is a diagram showing the power relationship between the source and load sides under 100 preset scenarios in a specific embodiment of the present invention when the high-voltage customer is not regulated.

[0074] Figure 7 This is a diagram showing the power relationship between the source and load sides when high-voltage customers adjust their power based on risk assessment results under 100 preset scenarios in a specific embodiment of the present invention. Detailed Implementation

[0075] This invention proposes a load control method based on external risk assessment of high-voltage customer electricity consumption. The invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0076] For ease of explanation, this embodiment first describes the five defined external risk assessment indicators for high-voltage customer electricity consumption:

[0077] Because distribution network faults and fluctuations in renewable energy output can cause phenomena such as node voltage fluctuations, power imbalances between source and load, line overloads, islanded nodes, and changes in node voltage harmonics, these phenomena can have varying degrees of impact on the safe electricity use of high-voltage customers. To ensure that high-voltage customers can reliably and safely use electricity when changes occur in the distribution network, five assessment indicators have been defined to form an external risk assessment indicator system for the safe electricity use of high-voltage customers.

[0078] 1) Node voltage fluctuation index

[0079]

[0080] Where: num1 is the number of devices inside the high-voltage customer that are sensitive to voltage fluctuations, num is the total number of devices inside the high-voltage customer, and p(C i ) is the preset scenario C i The probability value, U i U represents the actual voltage value at node i. Ni U is the rated voltage value of node i. imin U is the lower limit value of node i. imax Let i be the upper limit value of node i.

[0081] 2) Power imbalance index

[0082] X2=p(C i )P c (16)

[0083] In the formula: P c This will actually reduce the load for high-voltage customers.

[0084] 3) Line overload index

[0085]

[0086] In the formula: P j P represents the overload power of the j-th line connected to the node where the high-voltage customer is located. jNLet be the rated power of the j-th line, and k be the total number of lines connected to the node where the high-voltage customer is located.

[0087] 4) Isolated Node Indicators

[0088]

[0089] Where: ε j This is a criterion for determining whether node j, where the high-voltage customer is located, is an isolated node.

[0090] 5) Node voltage harmonic index

[0091]

[0092] In the formula: num2 represents the number of harmonic-sensitive devices within the high-voltage customer's premises, U j The effective value of the j-th harmonic voltage at the node where the high-voltage customer is located, the effective value of the fundamental voltage U1, and THD. u denoted as node voltage distortion rate.

[0093] Figure 1 The flowchart of the load control method based on external risk assessment of high-voltage customer electricity consumption according to the present invention includes the following steps:

[0094] Step A: Obtain the subjective, objective, and independence weights of the indicator system and construct an indicator classification set based on the independence weight distribution;

[0095] Step B: Establish the constraints and objective function of the self-learning integrated weight model based on the independence weight criterion;

[0096] Step C: Establish an evaluation and rating system that combines a self-learning comprehensive weight model with a cloud model;

[0097] Step D: Develop load control strategies based on external risk assessments of safe electricity use for high-voltage customers.

[0098] The specific steps in step A are as follows:

[0099] A1. Dataset of external risk indicators for high-voltage customer electricity consumption based on IEEE-24-RTS distribution network

[0100] Figure 2 The topology diagram of the IEEE-24-RTS distribution network is shown. The system state fast sorting method considering multi-state element model is used to obtain 8760 different preset scenarios of the IEEE-24-RTS distribution network and their probability values. Due to space limitations, only the top 100 preset scenarios with the highest probability values ​​are taken as examples, as shown in Table 1.

[0101] Table 1

[0102]

[0103]

[0104]

[0105]

[0106] Through power flow calculation, an electrical dataset of 8760 different preset scenarios with five combinations of external risk indicators for high-voltage customers was obtained, as shown in Table 2, providing a data foundation for the external risk assessment indicator system for high-voltage customers.

[0107] Table 2

[0108]

[0109]

[0110]

[0111]

[0112] A2. Obtain the subjective, objective, and independent weight matrix of the indicator system.

[0113] Based on the electrical dataset in Table 2, the subjective weights, objective weights, and independence weights corresponding to the five indicators were calculated using the AHP interval estimation optimization method, entropy method, and independent weight method. The results are shown in Table 3.

[0114] Table 3

[0115] Indicator 1 0.20157 0.35890 0.07625 Indicator 2 0.49302 0.05107 0.11597 Indicator 3 0.03593 0.06439 0.11609 Indicator 4 0.06792 0.48220 0.07625 Indicator 5 0.20157 0.04325 0.61544

[0116] A3. Indicator classification based on independence weights

[0117] Based on the independent weight values ​​in Table 3, the five indicators are classified according to their relative size relationship, and corresponding label sets are constructed, as shown in Table 4.

[0118] Table 4

[0119] Indicator 1 2 Reduced type Indicator 2 3 Uncertainty Indicator 3 1 Additive Indicator 4 2 Reduced type Indicator 5 1 Additive

[0120] Step B involves the following steps:

[0121] B1. Constraints of the self-learning integrated weight model based on the independence weight criterion

[0122] This specific implementation method involves three methods—subjective weighting, objective weighting—to combine weights, and assigns weights to five indicators, constructing a weight matrix A:

[0123]

[0124] The range of combined weights can be determined from the weight matrix A: in:

[0125]

[0126] The combined weight range is shown in Table 5.

[0127] Table 5

[0128] Indicator 1 0.04158 0.52899 Indicator 2 0.04014 0.49302 Indicator 3 0.03593 0.09520 Indicator 4 0.04159 0.23680 Indicator 5 0.06374 0.83655

[0129] Based on the above formula, construct the subjective interval constraint:

[0130]

[0131] Based on a label set H = {h1, h2, ..., h5} with 5 indicators, construct constraints according to the label types of the indicators:

[0132]

[0133] B2. Objective function of self-learning integrated weight model based on independence weight criterion

[0134] Different single weighting methods are used to determine the self-learning combined weights of subjective and objective weights:

[0135]

[0136] In the formula: the self-learning combined weight of the j-th indicator is β j1 β is the combination coefficient of the subjective weights of the j-th indicator. j2 is the combination coefficient of the objective weight of the j-th indicator.

[0137] The objective weights determined by the independence weight method reflect the relative proportions of each indicator when the overlap of information is minimized. Therefore, the relative unit slope d of the independence weights among each indicator is defined. i Reflecting the relative proportions among the various indicators:

[0138]

[0139] Using independence weights relative to the unit slope k i For reference, construct the objective function:

[0140]

[0141] In one specific embodiment, step C of the load control method based on self-learning high-voltage customer power consumption external risk assessment of the present invention includes the following specific steps:

[0142] C1. Determining the risk assessment matrix based on the cloud model

[0143] The cloud model is used to classify and quantify the external risk assessment index system for high-voltage customers' electricity use, and the risk level is divided into 4 levels. The specific index level range is divided and the cloud model parameters are determined.

[0144] Substitute the nonnegatively normalized indicator dataset into the following formula to calculate the membership degree R = [r1, r2, r3, r4] of each indicator to different risk levels.

[0145]

[0146] By integrating the membership matrices of each indicator, a risk assessment matrix R is constructed. ∑ :

[0147]

[0148] C2. A rating system combining a self-learning comprehensive weight model and a cloud model risk assessment matrix.

[0149] First, the self-learning combined weights based on the independence objective weight criterion are defined as the indicator layer weight vector W. R Based on the risk assessment matrix R determined by the cloud model, the criterion-level risk assessment matrix G, consisting of the risk assessment results of 100 preset scenarios, is calculated. ∑ And use it as the risk assessment matrix at the criterion level. Then, based on different preset scenario probability sets p(C i The summation and normalization are performed to obtain the criterion layer weight vector W. p ={p * (C1),p * (C2),…,p * (C m Combined with the risk assessment matrix G at the criterion level ∑ The target layer evaluation result Z = {z1, z2, z3, z4} is calculated. Based on the principle of maximum membership, the maximum value z in Z is selected. max The corresponding risk level is used to assess and classify the overall external risks of high-voltage customers' electricity use. The specific calculation formula is as follows.

[0150]

[0151] The corresponding weights are shown in Table 6.

[0152] Table 6

[0153]

[0154] The specific steps of step D are as follows:

[0155] D1. Screening power distribution network source-load imbalance scenarios and obtaining risk assessment results.

[0156] Based on A1, a fast sorting method considering the system state of a multi-state component model is used to obtain different preset scenarios and their probability values ​​for the IEEE-24-RTS distribution network. The top 100 preset scenarios with the highest probability values ​​are selected as examples. Through power flow calculations, a dataset X of power values ​​on both the source and load sides under 100 different preset scenarios is obtained. 100|2 ={X1,X2,…,X 100}, where X j ={P f ,P g}, P f P is the total power required by the distribution network load. g The total power that the generators in the distribution network can provide is shown in Table 7.

[0157] Table 7

[0158]

[0159]

[0160]

[0161]

[0162] Based on equation (20), the source-load imbalance scenario of the distribution network is screened, and a scenario label set Q = {a1, a2, ..., a...} is formed. 100}:

[0163]

[0164] Based on the scene label set Q = {a1, a2, ..., a...} 100}, select the scenes with a value of 1 to form the source-load imbalance scene set Q. s ={a1,a2,…,a s There are a total of s source-load imbalance scenarios. For each source-load imbalance scenario, the self-learning risk assessment method proposed in steps A to C is used to conduct external risk assessments for safe electricity use of all high-voltage customers connected to the distribution network, forming a risk set U = {b1, b2, ..., b...}. s}, where b1 represents the external risk level of safe electricity use for u high-voltage customers in scenario 1 of source-load imbalance, b1 = {b 11 ,b 12 ,…,b 1u In this embodiment, scenario 1 is taken as the analysis object, and the processing methods for other scenarios are similar to those for scenario 1.

[0165] D2. Formulate load control strategies based on risk assessment results.

[0166] D2.1, Adjustable high-voltage customer set based on risk set partitioning

[0167] According to Table 8, the adjustable risk level threshold θ = 2 is set in this example.

[0168] Table 8

[0169]

[0170]

[0171] Based on the risk set b1 of scenario 1 obtained from D1 (see Table 8), and compared with the set adjustable risk level threshold θ = 2, the adjustable high-pressure customer set K = {k1, k2, ..., k s See Table 9.

[0172] Table 9

[0173]

[0174]

[0175] Taking the first scenario as an example, the specific process of acquiring a set of adjustable high-voltage customers is explained:

[0176]

[0177] Where, k ji A value of 1 indicates that the high-voltage customer is controllable, while a value of 0 indicates that the high-voltage customer is not controllable.

[0178] D2.2 Load Reduction Strategy Model Based on Adjustable High-Voltage Customer Set

[0179] The load reduction strategy model for adjustable high-voltage customer sets takes the minimum load reduction amount as the objective function;

[0180]

[0181] Based on D1.1, K = {k1,k2,…,k} s The constraints of the load reduction optimization model for AC power flow and the constraints of the load reduction strategy model for an adjustable high-voltage customer set are determined.

[0182]

[0183] Among them, P t (V,δ)=V t ∑V r (G tr cosδ tr +B tr sinδtr );Q t (V,δ)=V t ∑V r (G tr cosδ tr -B tr sinδ tr );G th and B th δ represents the real and imaginary parts of the admittance matrix in row t and column h; V is the magnitude of the node bus voltage; δ is the phase angle difference between the two ends of the line; V t It is the voltage magnitude at node t; V r It is the voltage magnitude at node r; δ tr It is the phase angle difference between nodes t and r; C t P is the load reduction at node t; LDt Q represents the active load at node t; LDt The reactive load at node t; and These are the upper and lower limits of injected active and injected reactive power at generator node t, respectively; TR h It is the actual transmission capacity of the transmission line h; Yan is the maximum transmission capacity of the transmission line h; V t max and V t min These are the upper and lower limits of the voltage amplitude of the bus at node t, respectively; G, NG, N, L, S and U are the sets of load bus nodes, generator bus nodes, all bus nodes, all transmission lines, source-load imbalance scenarios and adjustable high-voltage customers in the power transmission system, respectively.

[0184] Based on the load reduction strategy model for adjustable high-voltage customers, and according to the specific adjustable high-voltage customer dataset in Table 9, the model is solved to obtain the required load shedding amount for each high-voltage customer. The specific values ​​are shown in Table 10.

[0185] Table 10

[0186]

[0187]

[0188] The specific analysis based on the data obtained above is as follows:

[0189] Through Table 8, Figure 3 , Figure 4It can be seen that the relative distribution of the weight values ​​of each indicator combination obtained in this embodiment is closer to the independent weight distribution than the relative distribution of the weight values ​​of the indicator combination obtained by the subjective and objective weighting combination evaluation method based on the product method. Moreover, considering that the independent weight distribution reflects the weight value under the condition of minimum information overlap between each indicator, this shows that the method proposed in this invention has a certain degree of ability to reduce the degree of information overlap between each indicator, making the weight of each indicator combination more reasonable.

[0190] pass Figure 5 It can be seen that: this embodiment and the product-based combination evaluation method, based on the principle of maximum membership, both determine the external risk assessment level of high-voltage customers to be Level 1, indicating that the method proposed in this invention is reasonable and effective; moreover, since the independence weight is used as a criterion to guide the combination of subjective and objective weights, the information redundancy of each indicator is reduced to a certain extent.

[0191] Tables 8, 9, and 10 show that the load reduction strategy model based on the adjustable high-voltage customer dataset can determine whether a customer is suitable for regulation based on the risk assessment results, from the perspective of ensuring the operational safety of high-voltage customers. Table 10 shows that only high-voltage customers with loads below the risk threshold θ = 2 are regulated to varying degrees, while those above the risk threshold θ = 2 are regulated by 0 MW. Furthermore, the comparison results in Table 11 show that before regulation, generators could not provide the required power to the load, while after regulation, generator output met the total power demand of the load. Therefore, the load regulation method proposed in this invention, based on external risk assessment of high-voltage customer electricity consumption, not only ensures the safe electricity consumption needs of the regulated high-voltage customers but also satisfies the purpose of source-load balance in the distribution network.

[0192] Table 11

[0193]

[0194] Meanwhile, to illustrate the universality of the method proposed in this invention, in addition to a careful analysis of scenario 1, the same method was used to reduce load in 100 preset scenarios with power imbalance between the source and load sides: scenarios 36, 41, 46, 50, 52, 57, 60, 70, 73, 75, 86, 90, 93, 95, and 96. The comparison of load and total generator power before and after regulation of high-voltage customers in each scenario is shown in Table 12.

[0195] Table 12

[0196]

[0197]

[0198] To more intuitively demonstrate the effect of the load control method based on external risk assessment of high-voltage customer electricity consumption proposed in this invention, through... Figure 6 , 7 The comparison shows that the method proposed in this invention achieves power balance between the source and load sides while ensuring the operational safety of the regulated high-voltage customer.

Claims

1. A load regulation method based on high-voltage customer electricity external risk assessment, characterized in that, The method includes the following steps: Step A: Obtain the dataset of external risk indicators for high-voltage customers' electricity consumption, as well as the subjective, objective, and independence weights of the indicator system, and construct an indicator classification set based on the distribution of independence weights; Step B: Establish the constraints and objective function of the self-learning integrated weight model based on the independence weight criterion; The constraints in step B include: Subjective interval constraints: ， Indicator label type constraints: ， In the formula: , Indicators Upper and lower limits of the weighting interval; The objective function in step B is: ， In the formula: For the first Self-learning combined weights of each indicator, For the first The combination coefficient of subjective weights of each indicator. For the first The combination coefficient of the objective weights of each indicator. The relative unit slope represents the independence weights among the indicators; Step C: Determine the risk assessment matrix and establish an assessment and rating system that combines a self-learning comprehensive weight model with a cloud model; Step D: Develop a load control strategy based on an assessment of external risks associated with high-voltage customers' electricity consumption; Step D specifically includes the following steps: Step D1: Obtain power value datasets from both the source and load sides under 100 different preset scenarios. ,in , The total power required by the distribution network load. The total power that the generators in the distribution network can provide; Step D2: Based on equation (12), filter the power distribution network source-load imbalance scenarios and form a scenario tag set. : ， Step D3: Based on the scene tag set Filter out the scenes with a value of 1 to form a source-load imbalance scene set. The self-learning risk assessment method in steps A to C is used to conduct external risk assessments on the safe electricity use of all high-voltage customers connected to the distribution network, forming a risk set. ,in For u high-voltage customers in scenario 1 of source-load imbalance, the external risk level of safe electricity use is determined, and ; Step D4, based on risk set Divide the high-voltage customer base into adjustable categories; set different adjustable risk level thresholds based on actual control needs. : Step D5: Risk set obtained in step D1 The adjustable risk level threshold set in step D4 Comparison to obtain adjustable high-voltage customer set ; Step D6: Construct a load reduction strategy model based on the adjustable high-voltage customer set K; Step D7: Solve for the total load that each adjustable high-voltage customer needs to control based on the load reduction strategy model in step D6.

2. The load control method based on external risk assessment of high-voltage customer electricity consumption according to claim 1, characterized in that, Step A specifically includes the following sub-steps: Step A1: Obtain the distribution network using a fast sorting method that considers the system states of multi-state components. A set of different preset scenarios and their probability values; Obtain through power flow calculation Under different preset scenarios Electrical datasets with multiple index combinations ; Step A2: Using the analytic hierarchy process (AHP) and interval estimation optimization method, a discriminant matrix is ​​constructed based on expert experience. Combined with electrical datasets Obtain the subjective weight values ​​of each indicator. Estimated range of subjective weights for each indicator The subjective weight matrix of the external risk index system for high-voltage customer electricity consumption is obtained by combining these components. ; Step A3: Using the entropy method, based on the electrical dataset... Obtain the objective weight values ​​of each indicator. And combine them to obtain the objective weight matrix of the external risk index system for high-voltage customer electricity consumption. ; Step A4: Using the independence method, based on the electrical dataset... Obtain the independence weight values ​​of each indicator. The independent objective weight matrix of the external risk indicator system for high-voltage customer electricity consumption is obtained by combining the results. : Step A5, The weight value of each selected indicator is compared with the weight values ​​of the remaining unselected indicators, and the indicators are classified based on independent objective weights to obtain... Category label set of each indicator .

3. The load control method based on external risk assessment of high-voltage customer electricity consumption according to claim 2, characterized in that, In step A2, the estimation range of the subjective weights of each indicator is obtained. The formula is as follows: ， In the formula: For the first The subjective weight value of each indicator, The discriminant matrix of the analytic hierarchy process (AHP) is... for Subjective weight matrix of each indicator system; Obtain the subjective weight values ​​of each indicator. The formula is as follows: ， Where E is the maximum entropy value corresponding to the maximum entropy criterion.

4. The load control method based on external risk assessment of high-voltage customer electricity consumption according to claim 2, characterized in that, In step A3, the objective weight values ​​of each indicator are obtained. The formula is as follows: ， In the formula: The preset total number of scenes, For the first The first indicator The proportion of each preset scenario For the first The entropy value of each indicator, For the first Each indicator has an objective weight value, and when... hour, .

5. The load control method based on external risk assessment of high-voltage customer electricity consumption according to claim 2, characterized in that, In step A4, the independence weight values ​​of each indicator are obtained. The formula is as follows: ， In the formula: For the first The multiple correlation coefficient of each indicator, for The average value, for Eliminate the remaining matrix. For the first The independent objective weight of each indicator.

6. The load control method based on external risk assessment of high-voltage customer electricity consumption according to claim 2, characterized in that, In step A5, the acquisition Category label set of each indicator The formula is as follows: ， ， In the formula: For the first Individual indicators weight value Greater than the A set of preset scenario indicators The number of times the weight values ​​of the remaining indicators are displayed. For the first Individual indicators weight value Less than the A set of preset scenario indicators The number of times the weight values ​​of the remaining indicators are displayed. For the first Individual indicators weight value equal to the A set of preset scenario indicators The number of times the weight values ​​of the remaining indicators are counted; These correspond to the categories of increasing, decreasing, and uncertain, respectively.

7. The load control method based on external risk assessment of high-voltage customer electricity consumption according to claim 1, characterized in that, Step C specifically includes the following steps: Step C1: Substitute the non-negatively normalized indicator dataset into equation (10) to calculate the membership matrix of each indicator belonging to different risk levels. , ， Step C2: Calculate the membership matrix of each indicator. Integrate and construct a risk assessment matrix : ， In the formula: The total number of preset scenes, For the first Each indicator belongs to the risk level. The membership degree value; Step C3: Define the self-learning combined weights based on the independence objective weight criterion as the indicator layer weight vector. The risk assessment matrix determined by combining cloud models Calculation obtained by The criterion-level risk assessment matrix consists of the risk assessment results of each preset scenario. And use it as the risk assessment matrix at the criterion level; Step C4: Divide the probability sets of different preset scenarios Summation and normalization are performed to obtain the criterion layer weight vector. Combined with the risk assessment matrix at the criterion level The target layer evaluation results are calculated. ; Step C5: Based on the principle of maximum membership, ... The maximum value in The corresponding risk level is used to assess and classify the overall external risks of high-voltage customers' electricity consumption.

8. The load control method based on external risk assessment of high-voltage customer electricity consumption according to claim 1, characterized in that, The load reduction strategy model includes: Objective function; ， in This refers to the load reduction amount; Constraints: ， in, ; ; and Represents the admittance matrix of the first Line 1 The real and imaginary parts of a column; It is the amplitude of the node bus voltage; It is the phase angle difference between the two ends of the line; It is a node The voltage amplitude; It is a node The voltage amplitude; It is a node and Phase angle difference between the two ends; It is a node The amount of load reduction; For nodes Active load on; For nodes reactive load on; , , and These are generator nodes The upper and lower limits of injected active and injected reactive power; It is a power transmission line The actual transmission capacity; Strictly speaking, it is a power transmission line. Maximum transmission capacity; and These are nodes Upper and lower limits of bus voltage amplitude; , , , , and These are the load bus nodes, generator bus nodes, all bus nodes, all transmission lines, source-load imbalance scenarios, and controllable high-voltage customers in the power transmission system.

Images