CN110097223A - Early warning method for damage of power transmission line under typhoon disaster - Google Patents

Abstract

The invention provides a method for early warning damage of a power transmission line under typhoon disasters. The method comprises the steps of collecting meteorological information, equipment operation information, micro-terrain information and real-time damage information to construct a training set; information quantization, default value processing and standardization processing are carried out on the information in the training set; establishing a power transmission line damage probability hybrid prediction model based on extreme value I type probability distribution, a random forest method and a Monte Carlo method, and calculating damage probability; and calculating risk indexes according to the damage probability, and performing emergency repair after the disaster. The prediction model has stronger adaptability and higher reliability, and can reasonably estimate the needed emergency repair resources from the macroscopic view.

Description

A kind of early warning method for transmission line damage under typhoon disaster

technical field

The invention belongs to the field of power system risk assessment, and in particular relates to an early warning method for damage to power transmission lines under typhoon disasters.

Background technique

With the economic development, the power infrastructure of the power grid has been continuously improved, but natural disasters will cause serious threats to the power grid equipment. Among them, the threat of typhoon disasters to the power grid is the most significant. Especially in coastal areas, typhoon disasters have caused serious damage to power grid facilities. Accidents such as large-scale power outages have caused a great impact on economic development and people's production and life. Therefore, there is an urgent need to establish effective early warning means.

In the past, some methods of early warning of transmission line damage did not consider the influencing factors comprehensively, such as the micro-topographic factors that have been proved to have an important impact in engineering practice; although some methods considered more comprehensive factors, the calculation method was more complex and not suitable for large-scale Simulation calculation; some methods only use the deterministic wind speed for calculation, and fail to consider the uncertainty of the predicted wind speed, so the applicability is not strong.

It can be seen that these methods in the past have shortcomings such as insufficient data analysis considerations, insufficient applicability, and low efficiency of causal algorithms in the early warning of transmission line damage. Therefore, there is an urgent need to study a damage early warning method with comprehensive considerations, strong adaptability and high computational efficiency.

SUMMARY OF THE INVENTION

In order to solve the problems of insufficient consideration factors, weak applicability and low algorithm efficiency in the existing transmission line damage early warning method, the present invention proposes a transmission line damage early warning method under typhoon disaster.

The technical scheme of the present invention is an early warning method for the damage of power transmission lines under a typhoon disaster, which specifically comprises the following steps:

Step 1: Collect meteorological information, equipment operation information, micro-terrain information, and real-time damage information to construct a training set;

Step 2: Perform information quantification, default value processing, and standardization processing on the information in the training set;

Step 3: Establish a hybrid prediction model of transmission line damage probability based on extreme value I-type probability distribution, random forest method, and Monte Carlo method, and calculate the damage probability;

Step 4: Calculate the risk index according to the probability of damage, and carry out emergency repair after the disaster.

Preferably, the meteorological information in step 1 includes that the maximum gust is V;

The equipment operation information in step 1 includes that the operation time is T and the design wind speed is V';

The micro-topography information described in step 1 includes that the altitude is H, the slope aspect is A, the slope is S, the slope position is P, the underlying surface type is U, and the surface roughness is R;

The real-time damage information described in step 1 includes that the damage status of the equipment is Y;

The information in the training set described in step 1 is:

The maximum gust is V(m/s), the equipment running time is T(year), the design wind speed is V′(m/s), the altitude is H(m), the slope aspect is A(°), and the slope is S(°). ), the slope position is P, the underlying surface type is U, the surface roughness is R(m), and the damage state is Y; each data in the training set is a vector of length 10 composed of the above variables.

Preferably, the process of quantizing the information in the training set described in step 2 is:

Since V, T, V', H, A, S, P, U, and R are all data formats, only the information Y in the training set is quantized, and the damaged state after quantization is represented by Y*;

The default value processing for the information in the training set described in step 2 is:

The default values of V, T, V', H, A, S, P, U and R are filled with the median respectively, and the processed variables are:

After the default value processing, the maximum gust is V _f , the equipment running time after the default value processing is T _f , the design wind speed after the default value processing is V _f ′, the altitude after the default value processing is H _f , and the default value processing The slope aspect is A _f , the default value of the slope after processing is S _f , the default value of the slope position after processing is P _f , the type of the underlying surface after the default value processing is U _f , and the surface roughness after the default value processing is R _f ;

The data standardization processing on the data after the default value processing described in step 2 is as follows:

Calculate V _f , T _f , V _f ′, H _f , A _f , S _f , P _f , U _f , and R _f respectively according to the standardized variables, and the specific formula is:

X ^* = (xx _min )/(x _max -x _min )

In the formula, x is the value of V _f , T _f , V _f ′, H _f , A _f , S _f , P _f , U _f , and R _f in the data after the default value processing, and x _min is the default value processing The minimum value among the post-data V _f , T _f , V _f ′, H _f , A _f , S _f , P _f , U _f , and R _f , x _max is the default value of the post-processing data V _f , T _f , V The maximum value among _f ′, H _f , A _f , S _f , P _f , U _f , and R _f , X* is the variable after normalization, namely: the maximum gust after normalization is V*, and the running time of the equipment after normalization is T* , the design wind speed after standardization is V′*, the altitude after standardization is H*, the slope aspect after standardization is A*, the slope after standardization is S*, the slope position after standardization is P*, the type of underlying surface after standardization is U*, The surface roughness after normalization is R*;

Preferably, the process based on the extreme value type I probability distribution described in step 3 is as follows:

Parameter estimation is performed on the extreme value type I probability distribution. Assuming that V _f obeys the extreme value type I probability distribution, the calculation formula of the extreme value type I probability distribution function is:

In the formula, a is the scale parameter of the distribution; u is the location parameter of the distribution, and the calculation formula of the probability density distribution function is:

The scale parameter and the location parameter can be estimated using the method of moments estimation.

The first-order moment (mathematical expectation) is calculated as:

Among them: y≈0.57722, the second-order moment (variance) calculation formula is:

The calculation formula of the scale parameter a is thus obtained:

The calculation formula of the position parameter u is:

Then the probability of occurrence of the maximum wind speed x _P that occurs once in N years is the guarantee rate, and its calculation formula is:

P(x _p )=P(V _f >x _P )=1-P(V _f ≤x _p )=1-F(x _p )

In the formula, P(x _p ) is the probability of occurrence of the maximum wind speed once in N years;

Based on the extreme value type I probability distribution, the distribution of the wind field is simulated and predicted in units of wind speed points. The accuracy of the wind speed point can reach 1km × 1km. It is assumed that the maximum gust V _f at each gust point at different times satisfies the extreme value type I probability distribution. , use V _f to calculate a and u, fit an extreme value I-type distribution for each wind speed point, and realize the simulation of wind field distribution;

The process of random forest method described in step 3 is as follows:

The input to the random forest algorithm is:

V*, V′*, T*, H*, A*, S*, P*, U*, R*;

The output is:

The probability that the damaged state is Y*=1;

The specific process of the Monte Carlo method described in step 3 is as follows:

When determined based on the extreme value I-type probability distribution, the Monte Carlo method is used to randomly extract the wind speed, and in each random wind field, assuming that the wind speed of each wind speed point is relatively independent, the distribution of the entire wind field can be simplified as Gust distribution at a single wind speed point;

M times wind speeds are randomly generated according to uniform distribution at N wind speed points, and the jth wind speed at the ith wind speed point is defined as Vi _,j , where i=1,2,...,N is the sequence of wind speed points , j=1,2,...,M is a sequence of random samples, and at the same time, the probability of occurrence of each random wind speed P (V _f =V _i,j ) is calculated by using the fitted extreme value I-type probability distribution;

The calculated damage probability described in step 3 is:

In each random wind field, the RF method is used to calculate the damage probability f of the tower ( _xi |V _f =V _i,j ), where x _i is the eigenvector at the wind speed point i, and V _f is the wind speed of x _i component, according to the Monte Carlo method, the damage probability of the tower at this wind speed point is equivalent to the average value of the M prediction results, and the calculation formula of the damage probability of the tower at each wind speed point is:

In the formula: M is the total sampling times, and Vi _,j is the wind speed generated by the jth time at the wind speed point i;

Preferably, the steps of calculating the risk index described in step 4 are as follows:

The damage probability is set to 0.5 as the threshold value, P _i >0.5 means the tower is damaged, P _i <0.5 means the tower is not damaged, and the calculation formula of the average Num of equipment damage units is:

The formula for calculating the average repair time T _Re is:

The formula for calculating the average value C of the repair cost is:

The formula for calculating the average value L of manpower requirements is:

The calculation formula of the average value V of the demand for emergency repair vehicles is:

In the formula, t is the repair time per unit of equipment; c is the repair cost per unit; m is the manpower required per unit of equipment; v is the repair cost per unit; ) is the indicator function, which is defined as follows:

The post-disaster repairs described in Step 4 are:

After obtaining the damage probability Pi of each wind speed point, at the same time, the risk indicators Num, C, L, and _V can guide the disaster prevention and mitigation department to pre-allocate equipment, expenses, personnel, and vehicle resources, and release the power outage duration early warning information according to the T _Re indicator. .

The present invention has the following advantages:

Comprehensive consideration of meteorological information, equipment operation information, micro-topography information, traffic information, vegetation information, real-time damage information, etc., the factors considered are relatively comprehensive;

Using extreme value I-type probability distribution and Monte Carlo method to simulate the random wind field many times, considering the uncertainty of the predicted wind field, the hybrid prediction model of transmission line damage probability has stronger adaptability than using the deterministic predicted wind speed. and higher credibility;

The damage probability prediction model based on random forest method has high computational efficiency, especially suitable for large-scale prediction;

The calculation of risk indicators is based on historical statistical data and a mixed prediction model of transmission line damage probability, which can reasonably estimate the required emergency repair resources from a macro perspective.

Description of drawings

Fig. 1: method flow chart of the present invention;

Figure 2: Prediction of damage by a hybrid prediction model of damage probability in a wind farm.

Detailed ways

In order to facilitate the understanding and implementation of the present invention by those of ordinary skill in the art, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the embodiments described herein are only used to illustrate and explain the present invention, but not to limit it. this invention.

Below in conjunction with Fig. 1 to Fig. 2, the specific embodiment of the present invention is introduced, including the following steps:

Step 1, collecting meteorological information, equipment operation information, micro-topographic information, and real-time damage information to construct a training set;

The meteorological information described in the step 1 includes that the maximum gust is V, and ArcGIS10.4.1 is used to carry out inverse distance weight interpolation processing to the maximum gust V in the meteorological information, and the relevant data is extracted to the coordinate point where the equipment is located.

The equipment operation information in step 1 includes that the operation time is T and the design wind speed is V';

The micro-topographic information in step 1 includes the altitude as H, the slope aspect as A, the slope as S, the slope position as P, the underlying surface type as U, and the surface roughness as R. Distance weight interpolation processing, and extract relevant data to the coordinate point where the device is located;

The real-time damage information described in the step 1 includes that the damage state of the equipment is Y, and is provided by the remote monitoring or inspection personnel of the electric power department, and the present embodiment uses 1 to represent the tower damage state, and 0 represents the tower undamaged state;

The information in the training set described in step 1 is:

The maximum gust is V(m/s), the equipment running time is T(year), the design wind speed is V′(m/s), the altitude is H(m), the slope aspect is A(°), and the slope is S(°). ), the slope position is P, the underlying surface type is U, the surface roughness is R(m), and the damage state is Y; each data in the training set is a vector of length 10 composed of the above variables.

Step 2, performing information quantification, default value processing, and standardization processing on the information in the training set;

The process of quantifying the information in the training set described in step 2 is as follows:

Since V, T, V', H, A, S, P, U, and R are all data formats, only the information Y in the training set is quantized, and the damaged state after quantization is represented by Y*;

The default value processing for the information in the training set described in step 2 is:

The default values of V, T, V', H, A, S, P, U and R are filled with the median respectively, and the processed variables are:

After the default value processing, the maximum gust is V _f , the equipment running time after the default value processing is T _f , the design wind speed after the default value processing is V _f ′, the altitude after the default value processing is H _f , and the default value processing The slope aspect is A _f , the default value of the slope after processing is S _f , the default value of the slope position after processing is P _f , the type of the underlying surface after the default value processing is U _f , and the surface roughness after the default value processing is R _f ;

The data standardization processing on the data after the default value processing described in step 2 is as follows:

Calculate V _f , T _f , V _f ′, H _f , A _f , S _f , P _f , U _f , and R _f respectively according to the standardized variables, and the specific formula is:

X ^* = (xx _min )/(x _max -x _min )

In the formula, x is the value of V _f , T _f , V _f ′, H _f , A _f , S _f , P _f , U _f , and R _f in the data after default value processing, and x _min is the default value processing The minimum value among the post-data V _f , T _f , V _f ′, H _f , A _f , S _f , P _f , U _f , and R _f , x _max is the default value of the post-processing data V _f , T _f , V The maximum value among _f ′, H _f , A _f , S _f , P _f , U _f , and R _f , X* is the variable after normalization, namely: the maximum gust after normalization is V*, and the running time of the equipment after normalization is T* , the design wind speed after standardization is V′*, the altitude after standardization is H*, the slope aspect after standardization is A*, the slope after standardization is S*, the slope position after standardization is P*, the type of underlying surface after standardization is U*, The surface roughness after normalization is R*;

Step 3, based on the extreme value I-type probability distribution, random forest method, and Monte Carlo method, establish a mixed prediction model of transmission line damage probability, and calculate the damage probability;

The process based on the extreme value type I probability distribution described in step 3 is as follows:

Parameter estimation is performed on the extreme value type I probability distribution. Assuming that V _f obeys the extreme value type I probability distribution, the calculation formula of the extreme value type I probability distribution function is:

In the formula, a is the scale parameter of the distribution; u is the location parameter of the distribution, and the calculation formula of the probability density distribution function is:

The scale parameter and the location parameter can be estimated using the method of moments estimation.

The first-order moment (mathematical expectation) is calculated as:

Among them: y≈0.57722, the second-order moment (variance) calculation formula is:

The calculation formula of the scale parameter a is thus obtained:

The calculation formula of the position parameter u is:

Then the probability of occurrence of the maximum wind speed x _P that occurs once in N years is the guarantee rate, and its calculation formula is:

P(x _p )=P(V _f >x _P )=1-P(V _f ≤x _p )=1-F(x _p )

In the formula, P(x _p ) is the probability of occurrence of the maximum wind speed once in N years;

Using the hourly forecast data 24 hours before the typhoon landed, matching the wind speed point for each tower based on the nearest geographical distance, and using each tower as a new wind speed point, the accuracy of the wind speed point can reach 1km × 1km. Based on the extreme value type I probability distribution, the distribution of the predicted wind field is simulated in units of wind speed points. Assuming that the maximum gust V _f at each gust point at different times satisfies the extreme value type I probability distribution, use V _f to calculate a and u, as Each wind speed point is fitted with an extreme value I-type distribution to realize the simulation of wind field distribution;

The process of random forest method described in step 3 is as follows:

The input to the random forest algorithm is:

V*, V′*, T*, H*, A*, S*, P*, U*, R*;

The output is:

The probability that the damaged state is Y*=1;

The specific process of the Monte Carlo method described in step 3 is as follows:

When determined based on the extreme value I-type probability distribution, the Monte Carlo method is used to randomly extract the wind speed, and in each random wind field, assuming that the wind speed of each wind speed point is relatively independent, the distribution of the entire wind field can be simplified as For the gust distribution at a single wind speed point, since the wind speed point has been matched to the coordinates of the tower, the damage probability at the wind speed point is actually the damage probability of the transmission line;

M times wind speeds are randomly generated according to uniform distribution at N wind speed points, and the jth wind speed at the ith wind speed point is defined as Vi _,j , where i=1,2,...,N is the sequence of wind speed points , j=1,2,...,M is the sequence of random samples, set N=83973, M=50, and use the fitted extreme value I-type probability distribution to calculate the probability P(V _f =V _i,j );

The calculated damage probability described in step 3 is:

In each random wind field, the RF method is used to calculate the damage probability f of the tower ( _xi |V _f =V _i,j ), where x _i is the eigenvector at the wind speed point i, and V _f is the wind speed of x _i component, according to the Monte Carlo method, the damage probability of the tower at this wind speed point is equivalent to the average value of the M prediction results, and the calculation formula of the damage probability of the tower at each wind speed point is:

In the formula: M is the total sampling times, V _i,j is the wind speed generated at the jth time at the wind speed point i. Use ArcGIS10.4.1 to visualize the results, and obtain the damage prediction of the mixed prediction model of damage probability under the predicted wind field shown in Figure 2.

Step 4: Calculate the risk index according to the probability of damage, and carry out emergency repair after the disaster.

The steps for calculating the risk indicator described in step 4 are as follows:

The damage probability is set to 0.5 as the threshold value, P _i >0.5 means the tower is damaged, P _i <0.5 means the tower is not damaged, and the calculation formula of the average Num of equipment damage units is:

The formula for calculating the average repair time T _Re is:

The formula for calculating the average value C of the repair cost is:

The formula for calculating the average value L of manpower requirements is:

The calculation formula of the average value V of the demand for emergency repair vehicles is:

In the formula, t is the repair time per unit of equipment; c is the repair cost per unit; m is the manpower required per unit of equipment; v is the repair cost per unit; ) is the indicator function, which is defined as follows:

The post-disaster repairs described in Step 4 are:

Statistics t=0.001124, m=1.21231, v=0.18320, c=0.28165. After obtaining the damage probability Pi of each wind speed point, at the same time, the risk indicators Num, C, L, and _V can guide the disaster prevention and mitigation department to pre-allocate equipment, expenses, personnel, and vehicle resources, and release the power outage duration early warning information according to the T _Re indicator. , the risk index Num=40, T _Re =0.45, M=49, V=8, C=11.266 are obtained in this example.

It should be understood that the above description of the preferred embodiments is relatively detailed, and therefore should not be considered as a limitation on the protection scope of the patent of the present invention. In the case of the protection scope, substitutions or deformations can also be made, which all fall within the protection scope of the present invention, and the claimed protection scope of the present invention shall be subject to the appended claims.

Claims (5)

1. transmission line of electricity damages method for early warning under a kind of typhoon disaster, which comprises the following steps:

Step 1: to weather information, equipment operation information, mima type microrelief information, damage information is acquired to construct training in real time Collection；

Step 2: information quantization, default value processing, standardization are carried out to information in training set；

Step 3: it is pre- that transmission line of electricity damage probability mixing being established based on extreme value type I probability distribution, random forest method, Monte Carlo Method Model is surveyed, and calculates damage probability；

Step 4: risk indicator calculating being carried out according to damage probability, is repaired after carrying out calamity.

2. transmission line of electricity damages method for early warning under typhoon disaster according to claim 1, which is characterized in that in step 1 The weather information includes that peak gust is V；

Equipment operation information described in step 1 include runing time be T, design wind speed is V '；

Mima type microrelief information described in step 1 include height above sea level be H, slope aspect A, gradient S, slope position be P, underlying surface type U, Table roughness is R；

Damaging the collapse state that information includes equipment described in step 1 in real time is Y；

Information in training set described in step 1 are as follows:

Peak gust is V (m/s), operation hours is T (year), design wind speed is V ' (m/s), height above sea level is H (m), slope aspect A (°), the gradient are S (°), slope position is P, underlying surface type U, roughness of ground surface are R (m), collapse state Y；It is every in training set Data is the vector that the length being made of above-mentioned variable is 10.

3. transmission line of electricity damages method for early warning under typhoon disaster according to claim 1, which is characterized in that in step 2 The process that information in training set is quantified are as follows:

Since V, T, V ', H, A, S, P, U, R are data format, only information Y in training set is quantified, shape is damaged after quantization State is indicated with Y*；

Default value processing is carried out to information in training set described in step 2 are as follows:

The default value of V, T, V ', H, A, S, P, U, R are filled using median respectively, the variable that obtains that treated are as follows:

Peak gust is V after default value processing_f, default value processing after operation hours be T_f, default value processing after design wind speed For V_f', default value processing after height above sea level be H_f, default value handle adverse grade to for A_f, default value processing after the gradient be S_f, at default value Reason adverse grade position is P_f, default value processing after underlying surface type be U_f, default value processing after roughness of ground surface be R_f；

Data normalization processing is carried out to data after default value processing described in step 2 are as follows:

Respectively by V_f、T_f、V_f′、H_f、A_f、S_f、P_f、U_f、R_fIt is calculated according to standardized variable, specific formula are as follows:

X^*=(x-x_min)/(x_max-x_min)

In formula, x is V in data after default value processing_f、T_f、V_f′、H_f、A_f、S_f、P_f、U_f、R_fIn value, x_minFor default value processing Data V afterwards_f、T_f、V_f′、H_f、A_f、S_f、P_f、U_f、R_fIn minimum value, x_maxFor data V after default value processing_f、T_f、V_f′、H_f、A_f、 S_f、P_f、U_f、R_fIn maximum value, X* be variable after standardization i.e.: peak gust be equipment fortune after V*, standardization after standardization The row time is T*, design wind speed is V ' * after standardization, height above sea level is H* after standardization, standardization adverse grade is to for A*, standardization adverse grade Degree is S*, standardization adverse grade position is P*, underlying surface type is U* after standardization, roughness of ground surface is R* after standardization.

4. transmission line of electricity damages method for early warning under typhoon disaster according to claim 1, which is characterized in that in step 3 The process based on extreme value type I probability distribution is as follows:

Parameter Estimation is carried out to extreme value type I probability distribution, it is assumed that V_fExtreme value type I probability distribution is obeyed, then extreme value type I probability distribution Function calculation formula are as follows:

In formula, a is the scale parameter of distribution；U is the location parameter of distribution, probability density function calculation formula are as follows:

Scale parameter and location parameter are estimated using moments estimation method；

First moment (mathematic expectaion) calculation formula are as follows:

Wherein: y ≈ 0.57722, second moment (variance) calculation formula are as follows:

Thus the calculation formula of scale parameter a is obtained are as follows:

The calculation formula of location parameter u are as follows:

The then maximum wind speed x that N mono- is met_PThe probability of appearance is fraction, its calculation formula is:

P(x_p)=P (V_f> x_P)=1-P (V_f≤x_p)=1-F (x_p)

In formula, P (x_p) it is the probability that N mono- meets that maximum wind speed occurs；

Based on extreme value type I probability distribution, using wind speed point as the distribution of unit simulation and forecast wind field, the precision of wind speed point be can reach 1km × 1km, it is assumed that peak gust V of different time at each fitful wind point_fMeet extreme value type I probability distribution, uses V_fA and u is calculated, It is fitted an extremum I distributing for each wind speed point, realizes wind field distribution simulation；

The process of random forest method described in step 3 is as follows:

The input of random forests algorithm are as follows:

V*,V′*,T*,H*,A*,S*,P*,U*,R*；

Output are as follows:

Collapse state is the probability of Y*=1；

Detailed process is as follows for Monte Carlo Method described in step 3:

After determining based on extreme value type I probability distribution, wind speed is randomly selected with Monte Carlo method, and in each RANDOM WIND FIELD Under, it is assumed that the wind speed of each wind speed point has relative independentability, then the distribution of entire wind field can be reduced at single wind speed point Fitful wind distribution；

According to M wind speed of generation at random is uniformly distributed on N number of wind speed point, defining jth time wind speed at i-th of wind speed point is V_i,j, In, i=1,2 ..., N are the sequence of wind speed point, and j=1,2 ..., M are the sequence of random sample, while utilizing the pole being fitted Value I type probability distribution calculates the probability P (V that each random wind speed occurs_f=V_i,j)；

The probability of calculating damage described in step 3 are as follows:

Under each RANDOM WIND FIELD, the damage probability f (x of shaft tower is calculated using RF method_i|V_f=V_i,j), wherein x_iFor wind speed point i The feature vector at place, V_fFor x_iWind speed component, according to Monte Carlo Method, shaft tower at wind speed point damage probability is equivalent to M times The average value of prediction result, the shaft tower at each wind speed point damage probability calculation formula are as follows:

In formula: M is total frequency in sampling, V_i,jThe wind speed generated for jth at wind speed point i time.

5. transmission line of electricity damages method for early warning under typhoon disaster according to claim 1, which is characterized in that in step 4 The step of risk indicator calculates is as follows:

Setting damage probability 0.5 is threshold value, P_i> 0.5 thinks that shaft tower is damaged, P_i< 0.5 expression shaft tower is not damaged, and equipment damages unit Several average value Num calculation formula are as follows:

The average value T of repair time_ReCalculation formula are as follows:

The average value C calculation formula of rehabilitation expense are as follows:

The average value L calculation formula of manpower demand are as follows:

The average value V calculation formula of recovery vehicle demand are as follows:

In formula, t is that unit equipment repairs duration；C is unit rehabilitation expense；M is manpower needed for unit equipment；V is unit reparation Expense；It is obtained by historical statistical data；[] is the function that rounds up, and I () is indicator function, is defined as follows:

It is repaired after progress calamity described in step 4 are as follows:

Obtain the damage probability P of each wind speed point_iAfterwards, while risk indicator Num, C, L, V can instruct the department of preventing and reducing natural disasters to be set The pre- allotment of standby, expense, personnel, vehicle resources, and according to T_ReIndex issues power failure duration warning information.