CN103995951A - Typhoon key parameter extraction method based on half-normal model - Google Patents

Abstract

The invention discloses a method for extracting key parameters of a typhoon based on a half-normal model. The invention inputs typhoon historical data into a computer, and the computer selects any point in the map data corresponding to a typhoon high-incidence area as a simulation point, and draws Simulate the circle to obtain the key parameters of the typhoon passing through the simulated circle; calculate the characteristic parameters of the half-normal model to obtain the characteristic parameters of the half-normal model of the key parameters of the typhoon of _Al and the characteristic parameters of the half-normal model of the key parameters of the typhoon of A _R. The invention has the characteristics of being able to accurately extract key parameters of typhoons, laying a reliable foundation for further research on typhoon intensity, typhoon path and impact of typhoon on people's lives, and providing reliable data basis for utilization of typhoon wind energy.

Description

Extraction method of key parameters of typhoon based on half-normal model

technical field

The present invention relates to the technical field of typhoon research, in particular to a key parameter of typhoon that can be accurately extracted, which lays a reliable foundation for further research on typhoon intensity, typhoon path and impact of typhoon on people's lives. Extraction Method.

Background technique

The key parameters of a typhoon directly determine the intensity of the typhoon, the path of the typhoon, and the impact of the typhoon on people's lives. The key parameters of a typhoon include central air pressure difference, moving speed, moving direction, etc. Usually, according to the frequency distribution diagram of the parameters, a certain probability distribution function (such as normal distribution, lognormal distribution, Weibull distribution, etc.) is used to simulate, and then the ^χ2 fitting test method and KS fitting test are used Finally, the probability model of key typhoon parameters at the simulation point is obtained, and the key parameters of typhoon are extracted according to the probability model of key typhoon parameters. However, the above methods cannot completely simulate the probability changes of key parameters, so the accuracy of the extracted key parameters is relatively poor.

Chinese Patent Authorization Publication No.: CN103177301A, authorized publication date June 26, 2013, discloses a typhoon disaster risk estimation method, which conducts statistical analysis on the loss data caused by typhoon disasters in designated monitoring areas, and selects the risk of disaster-causing factors The typhoon disaster risk assessment index system is based on the sensitivity of disaster-pregnant environment, the vulnerability of disaster-bearing bodies and the ability of disaster prevention and mitigation. The fuzzy transformation theory is used to establish a typhoon disaster risk prediction model, and the typhoon forecast results are used as the starting conditions of the prediction model. And input conditions, through the calculation and analysis of the prediction model, whether the predicted area will cause disasters in the future and the disaster risk level of disasters will be obtained, so as to improve the early warning ability of meteorological disasters. The weak point of this invention is, single function, can't extract typhoon key parameter.

Contents of the invention

The purpose of the invention is to overcome the deficiency of poor accuracy of key parameters that cannot be extracted by the methods in the prior art, and provide a key parameter that can accurately extract typhoons, so as to further study the typhoon intensity, typhoon path and typhoon’s impact on people’s lives The key parameter extraction method of typhoon based on half normal model has laid a reliable foundation.

A method for extracting key parameters of a typhoon based on a half-normal model, comprising the following steps:

(1-1) Input the typhoon historical data into the computer, the computer selects any point in the map data corresponding to the typhoon high-incidence area as a simulation point, takes the simulation point as the center, and R as the radius to make a simulation circle, and will pass through the simulation circle The central air pressure difference Δp of the typhoon, the typhoon moving direction, and the typhoon moving speed V _T are set as the key parameters of the typhoon at the simulation point;

The typhoon center pressure difference Δp refers to the difference between the typhoon center pressure and the typhoon peripheral pressure (generally 1010hPa), which is usually described by lognormal distribution and Weibull distribution;

The moving direction of the typhoon is calculated from the latitude and longitude of the center of the typhoon, which is usually described by a normal distribution or a double normal distribution;

The typhoon moving velocity V _T is obtained from the latitude and longitude of the record points before and after the typhoon center, which is usually described by normal distribution or lognormal distribution.

(1-2) Sample division:

computer formula Calculate the frequency of typhoon center pressure difference Δ _p , typhoon moving direction and typhoon moving speed V _T ;

Among them, x is any sample value of typhoon center pressure difference, typhoon moving direction or typhoon moving speed V _T , N is the total number of samples of typhoon center pressure difference, typhoon moving direction or typhoon moving speed V _T , f is typhoon center pressure Difference, frequency of typhoon moving direction or typhoon moving speed V _T ;

The computer draws the probability distribution histogram of typhoon center pressure difference Δp, typhoon moving direction and typhoon moving speed V _T ;

(1-2-1) Select the cut-off point of each probability distribution histogram:

For the histogram of the typhoon’s moving speed, the mean θ of the second distribution of the binormal distribution is selected as the first dividing point of the half-normal model; with the dividing point θ as the boundary, the moving direction of the typhoon is divided into the left half and the right half area; the peak value V _pmax of the histogram of the left area is used as the boundary point of the left probability distribution, and the peak value V _pmax of the histogram of the right area is used as the boundary point of the right probability distribution;

The computer selects the pressure difference Δp at the center of the typhoon and the peak value V _pmax of the typhoon movement direction histogram as the dividing point of the probability distribution;

(1-2-2) Perform the following processing on each probability distribution histogram:

Taking the boundary point as the boundary, divide the histogram of typhoon center pressure difference Δp and typhoon moving direction into left and right half areas, and obtain the left half area sample set A _l ={x _l |x≤V _pmax , x∈A } and the sample set A _r in the right half area = {x _r |x＞V _pmax , x∈A}; divide the typhoon movement speed histogram into four halves: the first left area, the second left area, the first right area, and the second right area district;

Among them, A is the sample set of pressure difference at the typhoon center, the sample set of typhoon moving direction or the sample set of typhoon moving velocity V _T , x _l is any sample value of A _l , x _r is any sample value of A _r ;

(1-3) Construct the symmetrical image sample set of each half area separately, and construct the half normal model:

Set A _l ′={x _l ′|x _l ′=2V _pmax -x _l , x _l ∈ A _l } as the symmetrical image sample set of A _l ＝{x _l |x≤V _pmax , x∈A}, A _l and A _l ′ form the first half-normal model;

Set A _r ′={x _r ′|x _r ′=2V _pmax -x _r , x _r ∈ A _r } as the symmetrical image sample set of A _r ＝{x _r |x＞V _pmax , x∈A}, A _r ′ and A _r form the second half normal model;

Set _AL as the left-half symmetric sample probability density set, and A _R the right-half symmetric sample probability density set; where, A _R ＝A _r ∪A _r ′, _AL ＝A _l ∪A _l ′;∪ is the union operator;

(1-4) Calculate the characteristic parameters of the half-normal model:

Using the parameter estimation method to calculate the expected value μ _L and _standard deviation δ _L of the wind speed data of AL, and the expected value μ _R and standard deviation δ _R of the wind speed data _of AR;

Obtain the characteristic parameters of the half-normal model of the typhoon key parameters of _AL The Characteristic Parameters of the Half-Normal Model of Typhoon Key Parameters in _AR

Among them, p is the proportionality coefficient, p=m _l /(m _l +m _r ), m _l is the number of samples of key parameters of typhoon in the left region, m _r is the number of samples of key parameters of typhoon in the right region, N represents the normal distribution, (p , 1] and [0, p] are the intervals of proportional coefficients.

The present invention firstly inputs typhoon historical data into the computer, and the computer selects any point in the map data corresponding to the typhoon high-incidence area as a simulation point, and makes a simulation circle, thereby obtaining key parameters of the typhoon passing through the simulation circle;

The computer draws a histogram of the probability distribution of typhoon center pressure difference Δp, typhoon moving direction and typhoon moving speed V _T ; with the dividing point as the boundary, the histogram of typhoon center pressure difference Δp and typhoon moving direction is divided into left and right half area, the left half area sample set A _l = {x _l |x≤V _pmax , x∈A} and the right half area sample set A _r ={x _r |x＞V _pmax , x∈A}; the typhoon The moving speed histogram is divided into four half-areas: the first area on the left, the second area on the left, the first area on the right, and the second area on the right; respectively construct a symmetrical image sample set for each half area, and construct a half-normal model; calculate the half-normal model Characteristic parameter, get the half-normal model characteristic parameter of the typhoon key parameter of _AL The Characteristic Parameters of the Half-Normal Model of Typhoon Key Parameters in _AR

The half-normal model of the invention can accurately reflect the ambiguity and correlation between key parameters, so that the key parameter rules can be extracted more accurately.

The typhoon historical data of the present invention comes from "CMA-STI Northwest Pacific Tropical Cyclone Best Track Dataset".

Preferably, R is 200 to 500KM.

As a preference, the moving direction of the typhoon is calculated from the latitude and longitude of the center of the typhoon, and four standard directions are set in the computer. The four standard directions are 0° to the north, 90° to the east, 180° to the south, and -90° to the west .

Preferably, the parameter estimation methods include least square method, maximum likelihood method, maximum posterior method, minimum risk method and minimization maximum entropy method.

Preferably, Δp is 0 to 135 hPa; typhoon moving speed V _T is 2 km/h to 65 km/h.

In order to achieve the above object, the present invention adopts the following technical solutions:

Therefore, the present invention has the following beneficial effects: (1) can accurately extract the key parameters of the typhoon, laying a reliable foundation for further research on typhoon intensity, typhoon path and typhoon's impact on people's lives; (2) provide reliable data for the utilization of typhoon wind energy Base.

Description of drawings

Fig. 1 is a kind of structural representation of simulation circle of the present invention;

Fig. 2 is a kind of flowchart of the embodiment of the present invention;

Figure 3 is a fitting diagram of typhoon movement direction by conventional method;

Fig. 4 is the fitting figure of typhoon moving direction in A district of the present invention;

Fig. 5 is the fitting diagram of typhoon movement direction in B area of the present invention;

Fig. 6 is a fitting diagram of typhoon moving direction in C zone of the present invention;

Fig. 7 is the fitting figure of typhoon movement direction in D area of the present invention;

Figure 8 is a fitting diagram of typhoon moving speed by conventional method;

Fig. 9 is a fitting diagram of typhoon moving speed in the left region of the present invention;

Fig. 10 is a fitting diagram of typhoon moving speed in the right region of the present invention;

Figure 11 is a fitting diagram of the typhoon center air pressure difference by the conventional method;

Fig. 12 is a fitting diagram of the left regional typhoon center air pressure difference of the present invention;

Fig. 13 is a fitting diagram of the air pressure difference at the center of the typhoon in the right region of the present invention.

In the figure: simulation point 1, simulation circle 2, coastline 3, typhoon track 4.

Detailed ways

The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

The embodiment shown in Figure 2 is a method for extracting key parameters of a typhoon based on a half-normal model, comprising the following steps:

Step 100, input the typhoon history data into the computer, the computer selects any point in the map data corresponding to the typhoon high-incidence area as the simulation point 1, takes the simulation point as the center, and R as the radius to make the simulation circle 2, and will pass through the simulation circle The typhoon's central air pressure difference Δp, typhoon moving direction, and typhoon moving speed V _T are set as the typhoon key parameters of the simulation point;

As shown in Figure 1, in the present embodiment, the computer selects any point in the map data corresponding to the coastline 3 as the simulation point, takes the simulation point as the center, and R=250KM as the radius to do the simulation circle, and the typhoon path 4 is simulated The central air pressure difference Δp of each typhoon in the circle, the typhoon moving direction, and the typhoon moving speed V _T are set as the key parameters of the typhoon at the simulation point;

As shown in Figure 2, step 200, sample division:

computer formula Calculate the frequency of typhoon center pressure difference Δp, typhoon moving direction and typhoon moving speed V _T ;

Among them, x is any sample value of typhoon center pressure difference, typhoon moving direction or typhoon moving speed V _T , N is the total number of samples of typhoon center pressure difference, typhoon moving direction or typhoon moving speed V _T , f is typhoon center pressure Difference, frequency of typhoon moving direction or typhoon moving speed V _T ;

The computer draws the probability distribution histogram of typhoon center pressure difference Δp, typhoon moving direction and typhoon moving speed V _T ;

The computer selects the histogram of the typhoon’s moving speed and selects the mean θ of the second distribution of the double normal distribution as the first cut-off point of the half-normal model; with the cut-off point θ as the boundary, the direction of the typhoon’s movement is divided into the left half and Right half area; the peak value V _pmax of the histogram of the left area is used as the boundary point of the left probability distribution, and the peak value V _pmax of the histogram of the right area is used as the boundary point of the right probability distribution;

The computer selects the pressure difference Δp at the center of the typhoon and the peak value V _pmax of the typhoon movement direction histogram as the dividing point of the probability distribution;

The computer divides each graph into different regions according to different demarcation points. For example, divide the typhoon center pressure difference Δp and typhoon moving direction into left and right half regions, and obtain the left half region sample set A _l ={ x _l |x≤V _pmax , x∈A} and the right half-area sample set A _r ＝{x _r |x＞V _pmax , x∈A}; divide the typhoon velocity histogram into four half-areas;

Among them, A is the sample set of pressure difference at the typhoon center, the sample set of typhoon moving direction or the sample set of typhoon moving velocity V _T , x _l is any sample value of A _l , x _r is any sample value of A _r ;

Step 300, respectively constructing a symmetrical image sample set for each half region, and constructing a half normal model:

Set A _l ′={x _l ′|x _l ′=2V _pmax -x _l , x _l ∈ A _l } as the symmetrical image sample set of A _l ＝{x _l |x≤V _pmax , x∈A}, A _l and A _l ′ form the first half-normal model;

Set A _r ′={x _r ′|x _r ′=2V _pmax -x _r , x _r ∈ A _r } as the symmetrical image sample set of A _r ＝{x _r |x＞V _pmax , x∈A}, A _r ′ and A _r form the second half normal model;

Set _AL as the probability density set of symmetric samples in the left half area, and the symmetric sample probability density set in the right half area of A _R ; among them, A _R ＝A _r ∪A _r ′, _AL ＝A _l ∪A _l ′;

Step 400, calculating the characteristic parameters of the half-normal model:

Using the parameter estimation method to calculate the expected value μ _L and _standard deviation δ _L of the wind speed data of AL, and the expected value μ _R and standard deviation δ _R of the wind speed data _of AR;

Obtain the characteristic parameters of the half-normal model of the typhoon key parameters of _AL The Characteristic Parameters of the Half-Normal Model of Typhoon Key Parameters in _AR

Among them, p is the proportional coefficient, p=m _l /(m _l + m _r ), m _l is the number of samples of key typhoon parameters in the left region, m _r is the number of samples of key parameters of typhoon in the right region, and N represents the normal distribution, (p, 1] and [0, p] are intervals of proportional coefficients.

Simulation test:

1. Generate the typhoon’s central air pressure difference Δp sample set, typhoon moving direction sample set, and typhoon moving speed V _T sample set and the mixed semi-cloud wind speed sample set:

(1) Use the randn(1) function that comes with MATLAB to randomly generate a random number a;

(2) Determine the value of a, if 0≤a≤p, use the normrnd command that comes with MATLAB to generate Random numbers that are distributed and located in the value range of the left area; if p<a≤1, use the normrnd command that comes with MATLAB to generate A random number distributed and located in the value range of the right area;

(3) Repeat steps (1) and (2) to obtain N data, sort the N data from small to large and form a mixed semi-cloud wind speed sample set;

2. Verify the half-normal model of the typhoon’s center pressure difference Δp sample set, typhoon moving direction sample set, and typhoon moving speed V _T sample set:

(1) Calculate the residual value

Sort the sample values x i of the typhoon central pressure difference Δp sample set, typhoon moving direction sample set or typhoon moving speed _{V I sample} set from small to large, using the formula Solve the residual value of the generated mixed semi-cloud wind speed sample set;

Among them, xi _′ is the sample value of the mixed semi-cloud wind speed sample set corresponding to _xi .

(2) Goodness of fit test

The goodness of fit test is a statistical method used to test whether a certain theoretical distribution is consistent with the original data distribution.

${\mathrm{<span\; class="notranslate">\; R</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\frac{{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; w</span>}}}{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; 100</span>}<span\; class="notranslate">\; \%</span>$

In the formula, e _w is the residual value; is the mean value of each sample value x _i in the sample set; the larger the value of R ² , the better the model fitting effect.

3. Comparison of simulation results

In this embodiment, the key parameters of a typhoon at a simulated point (E111.83°, N21.58°) extracted by conventional methods and half-normal were compared.

1. Typhoon movement direction

The normal distribution and double normal distribution are used to fit the probability distribution of typhoon movement direction respectively, as shown in Figure 3.

From the fitting effect shown in Figure 3, it can be seen intuitively that the fitting effect of the binormal distribution model is better than that of the normal distribution. According to the probability histogram of typhoon moving direction, the distribution has bimodal characteristics, and regions can be divided at the two peaks and the trough between the two peaks. The frequency histogram is divided into A, B, C and D areas, and the symmetrical data of the four areas are respectively constructed, and the normal model is used to fit them. The fitting effect is shown in Figure 4, Figure 5, Figure 6, and Figure 7.

The digital characteristics of the corresponding half-normal model were obtained from the four normally distributed data parameters, and the wind speed data were randomly generated using the normal model, the double normal model and the four half-normal models. The results of the fitting degree calculation are shown in Table 1 .

Table 1 Comparison of fitting degree of typhoon movement direction of different models

Model normal distribution binormal distribution half normal Fit % 92.480 7 99.341 0 99.508 3

From the results in Table 1, the half-normal model has the highest fitting degree, which is 7.027 6% and 0.167 3% higher than the normal distribution and double normal distribution, respectively, indicating that the half-normal model has a good fit for the direction of typhoon movement Effect.

2. Typhoon speed

According to the statistical results, the probability distribution of typhoon speed has unimodal characteristics, and the normal distribution and lognormal distribution models are used to fit them respectively. The fitting effect is shown in Figure 8.

Since the typhoon speed probability distribution has a unimodal characteristic, the probability distribution is divided into left and right regions, and image samples of the left and right regions are constructed, and the normal model is used to fit them respectively. The fitting effect is shown in Figure 9 and Figure 10.

The digital characteristics of the corresponding half-normal model were obtained from the data parameters of two normal distributions, and the wind speed data were randomly generated using the normal model, the lognormal model and two half-normal models, and the fitting degree calculation results are shown in Table 2 Show.

Table 2 Comparison of typhoon movement speed fitting degree of different models

Model normal distribution lognormal distribution half normal Fit % 92.1175 96.1593 97.2501

From the results in Table 2, the half-normal model has the highest fitting degree, which is 5.132 6% and 1.090 8% higher than the normal distribution and the lognormal distribution, respectively, indicating that the half-normal model has a good fit for typhoon moving speed. combined effect.

3. Air pressure difference at the typhoon center

According to the statistical results, the probability distribution of pressure difference in the center of the typhoon has a unimodal characteristic, and the lognormal distribution and Weibull distribution models are used to fit them respectively. The fitting results are shown in Figure 11.

Since the probability distribution of pressure difference in the typhoon center has a unimodal characteristic, the probability distribution is divided into left and right regions, and the image samples of the left and right regions are constructed, and the normal model is used to fit them respectively. The fitting effect is shown in Figure 12 and Figure 13.

The digital characteristics of the corresponding half-normal model were obtained from two normally distributed data parameters, and the lognormal model, Weibull model, and two half-normal models were used to randomly generate wind speed data. The fitting results are shown in Table 3 .

Table 3 Comparison of fitting degree of typhoon center pressure difference of different models

From the results in Table 3, the half-normal model has the highest fitting degree, which is 12.455 1% and 0.029% higher than the lognormal distribution and Weibull distribution, respectively, indicating that the half-normal model has a good fit for the typhoon center pressure difference Effect.

From the comparison of the calculation results of the above models, it can be seen that after the data probability distribution is divided into regions, the normal distribution is used to fit them separately, and after the parameters of the half-normal distribution are obtained, the data fitting of the original data has a high degree of fitting and versatility Therefore, the accuracy of the key parameters of the typhoon extracted is higher, laying a reliable foundation for further research on typhoon intensity, typhoon track and typhoon’s impact on people’s lives; providing a reliable data basis for the utilization of typhoon wind energy.

It should be understood that this embodiment is only used to illustrate the present invention but not to limit the scope of the present invention. In addition, it should be understood that after reading the teachings of the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms also fall within the scope defined by the appended claims of the present application.

Claims (5)

1. the typhoon key parameter extracting method based on half normal model, is characterized in that, comprises the steps:

(1-1) typhoon historical data is input in computing machine, any point in the computer selecting map datum corresponding with typhoon region occurred frequently is simulation points (1), taking simulation points as the center of circle, R is that radius does simulation circle (2), by central gas pressure reduction Δ p, Typhoon Tracks direction, the Typhoon Tracks speed V of the typhoon through simulation circle _tbe made as the typhoon crux parameter of described simulation points;

(1-2) sample is divided:

Computing machine formula calculate the poor Δ p of typhoon central pressure, Typhoon Tracks direction and Typhoon Tracks speed V _tfrequency;

Wherein, x is center of typhoon draught head, Typhoon Tracks direction or Typhoon Tracks speed V _tany sample value, N is center of typhoon draught head, Typhoon Tracks direction or Typhoon Tracks speed V _ttotal sample number amount, f is center of typhoon draught head, Typhoon Tracks direction or Typhoon Tracks speed V _tfrequency;

Computing machine draws center of typhoon draught head Δ p, Typhoon Tracks direction and Typhoon Tracks speed V _tprobability distribution histogram;

(1-2-1) select the histogrammic separation of each probability distribution:

The average θ of second distribution of the two normal distributions of histogram selection of Typhoon Tracks speed is as first separation of half normal model; Taking separation θ as boundary, Typhoon Tracks direction is divided into left half-court and right half-court; By histogrammic left side district peak value v _pmaxas the separation of left side probability distribution, by histogrammic the right district peak value v _pmaxas the separation of the right probability distribution;

Computing machine is selected the peak value v of center of typhoon draught head Δ p, Typhoon Tracks direction histogram _pmaxas the separation of probability distribution;

(1-2-2) each probability distribution histogram is all carried out to following processing:

Taking separation as boundary, the histogram of center of typhoon draught head Δ p, Typhoon Tracks direction is divided into left and right Liang Geban district, obtain left half-court sample set A _l={ x _l| x≤V _pmax, x ∈ A} and right half-court sample set A _r={ x _r| x > V _pmax, x ∈ A}; Typhoon Tracks velocity histogram is divided into the first from left district, the second from left district, You Yiqu, Si Geban district of You Er district;

Wherein, A is center of typhoon draught head sample set, Typhoon Tracks direction sample set or Typhoon Tracks speed V _tsample set, X _lfor A _lany sample value, x _rfor A _rany sample value;

(1-3) construct respectively the symmetry reflection sample set in each halfth district, structure obtains half normal model:

Set A _l'={ x _l' | x _l'=2V _pmax-x _l, x _l∈ A _lbe A _l={ x _l| x≤V _pmax, the symmetry reflection sample set of x ∈ A}, A _lwith A _l' composition the first half normal models;

Set A _r'={ x _r' | x _r'=2V _pmax-x _r, x _r∈ A _rbe A _r={ x _r| x > V _pmax, the symmetry reflection sample set of x ∈ A}, A _r' and A _rform the second half normal models;

Set A _lfor left half-court symmetrization sample probability density collection, A _rright half-court symmetrization sample probability density collection; Wherein, A _r=A _r∪ A _r', A _l=A _l∪ A _l';

(1-4) calculate half normal model characteristic parameter:

Utilize method for parameter estimation to calculate and obtain A _lthe expectation value μ of air speed data _lwith standard deviation δ _l, A _rthe expectation value μ of air speed data _rwith standard deviation δ _r;

Obtain A _lhalf normal model characteristic parameter of typhoon key parameter a _rhalf normal model characteristic parameter of typhoon key parameter

Wherein, p is scale-up factor, p=m _l/ (m _l+ m _r), m _lfor left region typhoon key parameter sample size, m _rfor right region typhoon key parameter sample size, N represents normal distribution, (p, 1] and [0, the p] interval that is scale-up factor.

2. the typhoon key parameter extracting method based on half normal model according to claim 1, is characterized in that, R is 200 to 500KM.

3. the typhoon key parameter extracting method based on half normal model according to claim 1, Typhoon Tracks direction is obtained by the position calculation of center of typhoon longitude and latitude, in computing machine, set 4 reference directions, 4 reference directions are respectively 0 ° of northern row, eastbound 90 °, 180 ° of southern row, head west-90 °.

4. the typhoon key parameter extracting method based on half normal model according to claim 1, is characterized in that, method for parameter estimation comprises least square method, maximum-likelihood method, Maximum Verified Method, minimum risk method and minimization Maximum entropy method.

5. according to the typhoon key parameter extracting method based on half normal model described in claim 1 or 2 or 3 or 4, it is characterized in that, Δ p is 0 to 135hPa; Typhoon Tracks speed V _tfor 2km/h to 65km/h.