Monte Carlo Markov chain-based strong typhoon wind speed and wind direction joint probability distribution generation method

Abstract

The invention discloses a Monte Carlo Markov chain-based strong typhoon wind speed and wind direction joint probability distribution generation method, which comprises the following steps of: firstly, simulating a typhoon wind field value through an established Emanuel typhoon radial section model, a near-surface typhoon wind speed conversion method, an average wind vertical section model and a fluctuating wind model; then, a typhoon whole process stochastic simulation method is used for carrying out stochastic simulation on the typhoon process, a simulation result is analyzed, the typhoon position at the next moment step is calculated, and on the basis, probability distribution calculation is carried out on a typhoon wind field to obtain a wind speed extreme value and a wind direction; and finally, constructing a combined probability model of the wind speed and the wind direction of the strong typhoon by adopting a two-dimensional Gaussian nonparametric kernel density estimation method. According to the method, a typhoon process stochastic simulation result and a typhoon wind field simulation method are combined, a more scientific and more practical solution is provided for wind resistance evaluation of the engineering structure, and key data support is provided for typhoon disaster vulnerability analysis of the engineering structure.

Description

Technical Field

[0001] The present invention relates to the technical field of wind disaster assessment, and to a method for generating a joint probability distribution of wind speed and wind direction in a strong typhoon landing area along the coast of my country. Specifically, a method is used to obtain the extreme value of typhoon wind speed and wind direction distribution in a target area by combining a typhoon process random simulation method with a typhoon wind field numerical simulation method, taking a strong typhoon as a basic object, and then generating a joint probability distribution of strong typhoon wind speed and wind direction. The method can be applied to the typhoon disaster vulnerability assessment of engineering buildings and structures. Background Art

[0002] Typhoons, as strong storms, are the main natural disasters that cause harm to my country's coastal areas, often leading to traffic suspension, damage to buildings and structures, and threats to personal safety. Therefore, in order to better prevent typhoon disasters, it is necessary to vigorously develop typhoon simulation technology and its probability assessment method.

[0003] At present, typhoon wind field analysis usually considers wind speed and wind direction separately to obtain their respective probability distributions. However, in the evaluation of wind resistance of engineering structures, it is often necessary to consider the combined effect of wind speed and wind direction. The joint probability distribution of the two is a key element in conducting typhoon disaster vulnerability analysis of engineering structures. However, the current method for obtaining this joint probability distribution is missing, which has become a key issue that needs to be solved in conducting typhoon disaster vulnerability analysis of engineering structures. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a method for generating a joint probability distribution of strong typhoon wind speed and wind direction based on a Monte Carlo Markov chain to solve the problems in the background technology.

[0005] To achieve the above object, the present invention provides the following technical solution: a method for generating a joint probability distribution of wind speed and wind direction of a strong typhoon based on a Monte Carlo Markov chain, comprising the following steps:

[0006] S1. First, the typhoon radial profile model of Emanuel typhoon and the near-surface typhoon wind speed conversion method, as well as the average wind vertical profile model and the pulsating wind model are established to achieve the numerical simulation of the typhoon wind field;

[0007] S2. Then, the typhoon process is stochastically simulated using the method of stochastic simulation of the entire process, including the initial simulation, progress simulation and termination simulation, to conduct stochastic simulation of the typhoon process and analyze the simulation results, and the typhoon position at the next time step is calculated by combining the simulation methods of each stage;

[0008] S3. Finally, the two-dimensional Gaussian nonparametric kernel density estimation method is used to calculate the probability distribution of the typhoon wind field to obtain the annual wind speed extreme value and wind direction data set, and to construct a joint probability model of strong typhoon wind speed and wind direction.

[0009] Preferably, in step S1, establishing the radial profile model of Typhoon Emanuel specifically includes:

[0010] Assuming that the air angular momentum and the partial derivative of angular momentum with respect to radius are continuous at the inner and outer junctions of the typhoon radial section, by introducing the continuity condition of air angular momentum at the inner and outer junctions, the typhoon full radial section model is finally obtained. The typhoon full radial section model includes three environmental parameters of the Emanuel typhoon radial section model, namely, the sea or surface resistance coefficient C D , the sinking rate w of the free tropospheric air in the outer area of ​​the typhoon and the ratio of the surface heat entropy exchange coefficient to the surface drag coefficient ξ, as well as 6 typhoon parameters, namely the maximum wind speed V of the typhoon m and its corresponding radius r m , the radius r of the intersection of the inner and outer regions a And the corresponding wind speed V a , typhoon scale r0 and latitude of typhoon center Then introduce the typhoon outer observation point (r * ,V * ), the internal and external joint solution method is used to obtain the full radial profile of the typhoon.

[0011] Preferably, in step S1, the near-surface typhoon wind speed is calculated, and the surface wind is divided into two components: one is the axisymmetric vortex wind V associated with the typhoon vortex r ; The second is the surface background wind V related to the local environment t ; The surface wind U is expressed as the weighted sum of the two components; the size of the vortex wind Vr is calculated by the Emanuel typhoon radial profile model, and the inflow angle α is introduced to correct its direction; the background wind V t The magnitude is equal to the typhoon's moving speed V t , and introduce a rotation angle β to correct its direction.

[0012] Preferably, in step S1, the vertical profile of the average wind speed is described by a logarithmic function, and the empirical formula of the vertical profile is corrected; according to the difference in average wind speed at different times and heights; the average wind speed U at time t at height z is equal to the product of the vertical profile of the average wind speed and the time function f(t); the time function f(t) is called the time profile of the average wind speed; U(z) is set to the maximum average wind speed value at height z during the typhoon history, then the time profile f(t) is a function less than or equal to 1.

[0013] Preferably, in step S1, the step of constructing a fluctuating wind model includes:

[0014] The fluctuating wind of a typhoon is assumed to be a Gaussian non-stationary process, and the evolutionary power spectrum density method is used to simulate the fluctuating wind, including assuming that the frequency composition of the fluctuating wind remains unchanged throughout the typhoon process, calculating the time-frequency conversion relationship and the evolutionary power spectrum density function of the non-stationary fluctuating wind speed, and then expressing the evolutionary power spectrum density of the non-stationary fluctuating wind by the modulation function and the power spectrum density of a certain stationary process. Then, the correlation of the fluctuating wind at two points in space along the vertical z and horizontal x is considered, and the coherence function proposed by Davenport is used. Finally, the evolutionary power spectrum density matrix of the non-stationary fluctuating wind speed of the typhoon is obtained in combination with the structural space examination points. Based on the above fluctuating wind spectrum, the harmonic synthesis method based on the superposition of trigonometric series is used to simulate the non-stationary process of the fluctuating wind. The Deodatis harmonic synthesis method is used to simulate the fluctuating wind field of the transmission tower-line system. After calculation, the Fourier transform technology is used to solve it. Finally, the evolutionary power spectrum of the non-stationary fluctuating wind time course is obtained.

[0015] Preferably, in step S2, the typhoon's initial simulation step includes: ① annual occurrence frequency sampling; ② based on the annual occurrence frequency, extracting the starting position information of each typhoon; ③ sequentially extracting the typhoon's initial intensity, initial moving speed, initial moving direction angle and initial scale information.

[0016] Preferably, in step S2, the typhoon progress simulation step includes: ① randomly extracting initial values ​​of various parameters of the typhoon based on the initial state probability model; ② randomly extracting and calculating the state change rate of the typhoon parameters for the next time step based on the state information of the typhoon at the current time step, combined with the state transition probability model corresponding to each parameter; ③ obtaining the typhoon state parameters and the typhoon center position for the next time step based on the state change rate sampling result; ④ saving the state information of the next time step as the current state of the typhoon, and then returning to step ② to perform the next step of typhoon state simulation.

[0017] Preferably, in step S2, the termination simulation adopts the partition statistics method to construct a spatially discrete typhoon termination probability model, and its specific steps include: ① defining the termination area range; ② dividing the grid within the termination area; ③ dividing the historical termination point samples into two categories: sea surface termination points and land termination points, and performing probability modeling respectively; ④ taking all historical typhoon path points as the object, counting the number of path points in each grid; ⑤ taking the historical termination points as the object, counting the number of termination points in each grid; ⑥ The ratio of the number of termination points to the number of path points is the typhoon termination probability of the corresponding grid; ⑦ If there is no historical path point in the grid, its termination probability is set to 1; if there is a historical path point but no historical termination point, the termination probability is 0.

[0018] Preferably, in step S3, the annual wind speed extreme value and its wind direction data set are calculated and obtained by the following steps:

[0019] 1) Based on the typhoon simulation results for 10,000 years, find the typhoon path points where the distance from the typhoon center point to the target location does not exceed the corresponding absolute scale of the typhoon, as well as their corresponding typhoon numbers and occurrence year numbers; the found path points, typhoon numbers, and occurrence year numbers are the objects for the next investigation;

[0020] 2) Count the typhoon numbers j (j = 1, 2,..., J) that affect the target location in the i-th year (i = 1, 2,..., I);

[0021] 3) Obtain the typhoon path points k (k = 1, 2,..., K) that affect the target location in the j-th typhoon;

[0022] 4) Based on the typhoon state parameters of the path point k, use the numerical simulation method of the typhoon wind field to calculate the surface wind field of the target area, that is, the wind speed and wind direction, and record them;

[0023] 5) If k < K, let k = k + 1 and return to step 4); if k = K, continue to the next step;

[0024] 6) Count the wind field data caused by all path points k (k = 1, 2,..., K) in the j-th typhoon on the target area, find the maximum wind speed value and its corresponding wind direction, and record them as the wind speed extreme value and the corresponding wind direction caused by the j-th typhoon;

[0025] 7) If j < J, let j = j + 1 and return to step (3); if j = J, continue to the next step;

[0026] 8) Count the wind speed extreme values and their wind directions caused by all typhoons j (j = 1, 2,..., J) in the i-th year, find the maximum wind speed extreme value and its corresponding wind direction, and record them as the annual wind speed extreme value and the corresponding wind direction caused by the typhoons in the i-th year;

[0027] 9) If i < I, let i = i + 1 and return to step (2); if i = I, continue to the next step;

[0028] 10) Count the annual wind speed extreme values and their wind directions of all year numbers i (i = 1, 2,..., I) and record them, and then construct a joint probability model of typhoon wind speed and wind direction in the target area.

[0029] Compared with the prior art, the beneficial effects of the present invention are:

[0030] 1. The present invention effectively solves the problem of missing joint probability distribution of wind speed and wind direction proposed in the background technology by combining multiple models, stochastic simulation and joint probability model construction. At the same time, the joint probability model constructed by the two-dimensional Gaussian nonparametric kernel density estimation method can more accurately reflect the combined effect of wind speed and wind direction, thereby providing a more scientific basis for the wind-resistant design of engineering structures. For example, in the typhoon probability vulnerability assessment of transmission lines, the failure probability of the line under different wind speed and wind direction combinations can be more accurately predicted, thereby optimizing the line design, reinforcement plan and emergency plan.

[0031] 2. The present invention uses the Emanuel typhoon radial profile model, the near-surface typhoon wind speed conversion method, the average wind vertical profile model and the pulsating wind model to achieve a multi-dimensional simulation of the typhoon wind field. Combined with the typhoon full-process random simulation method, it can more comprehensively reflect the dynamic change process of the typhoon and provide reliable basic data for subsequent probability distribution calculations and joint probability model construction.

[0032] 3. The present invention has broad application value in multiple fields of typhoon disaster prevention, promotes the development of typhoon simulation technology and probability assessment methods, provides a more scientific and practical solution for the evaluation of wind resistance of engineering structures, and provides key data support for the typhoon disaster vulnerability analysis of engineering structures. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0034] In the attached picture:

[0035] Figure 1 It is a schematic diagram of kernel density estimation; kernel density estimation is a non-parametric method for estimating probability density functions, which obtains a continuous probability density function by smoothing the histogram of a data set;

[0036] Figure 2 is the maximum wind speed V m distribution, Coriolis coefficient f distribution, typhoon moving speed V t Distribution and distribution of typhoon moving direction angle θ The overall distribution of four typhoon key parameters, including a comparative diagram of generalized extreme value distribution and kernel density estimation;

[0037] Figure 3 is the typhoon-related scale r * Distribution and typhoon absolute scale r0 distribution, including a comparative diagram of generalized extreme value distribution and kernel density estimation;

[0038] Figure 4It is the annual frequency distribution of typhoons, including a comparative diagram of Poisson distribution and binomial distribution;

[0039] Figure 5 It is a schematic diagram of the location information of the typhoon starting point and the fitting results using the two-dimensional non-parametric Gaussian kernel density estimation method;

[0040] Figure 6 It is a schematic diagram of the marginal probability density of typhoon intensity (left) and the joint probability density of intensity and its change rate (right);

[0041] Figure 7 It is a schematic diagram of the marginal probability density of the typhoon moving speed (left) and the joint probability density of the moving speed and its rate of change (right);

[0042] Figure 8 It is a schematic diagram of the marginal probability density of the typhoon moving direction angle (left) and the joint probability density of the moving direction angle and its change rate (right);

[0043] Fig. 9 Schematic diagram of the marginal probability density of the absolute scale of a typhoon (left) and the joint probability density of the absolute scale and its rate of change (right);

[0044] Fig.10 It is a schematic diagram of the typhoon termination simulation probability model and the fitting results of the termination point location distribution using the two-dimensional non-parametric Gaussian kernel density estimation method;

[0045] Fig.11 This is a flow chart of the Monte Carlo simulation of the entire typhoon path over 10,000 years;

[0046] Fig.12 It is a schematic diagram of the surface winds caused by typhoons in the Northern Hemisphere;

[0047] Fig.13 This is a schematic diagram of the extreme wind speed and wind direction distribution of typhoons in the sample area;

[0048] Fig.14 It is a schematic diagram of the annual extreme value probability distribution of typhoon wind field in the sample area;

[0049] Fig.15 It is a schematic diagram of the joint probability density of the annual extreme typhoon wind speed and its wind direction in the example area. DETAILED DESCRIPTION

[0050] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0051] Embodiment: A method for generating a joint probability distribution of wind speed and wind direction of a strong typhoon based on a Monte Carlo Markov chain, the specific steps include numerical simulation of a typhoon wind field, random simulation of a typhoon process, and probability distribution of a typhoon wind field; wherein the overall probability distribution of typhoon parameters based on historical data is used as a basis, and the path prediction based on the Markov chain theory is used as the core, combined with Monte Carlo sampling, to obtain the entire process of typhoon events in the northwest Pacific region for thousands of years. On this basis, all typhoon processes affecting the target inspection area are judged and extracted, and the Emanuel typhoon radial profile model and the near-surface typhoon wind speed conversion method are used to calculate and obtain the wind speed and wind direction data set of the target area, and the annual occurrence rate is used as a benchmark, combined with two-dimensional Gaussian non-parametric kernel density estimation, to generate a joint probability distribution model of the annual wind speed extreme value and wind direction of the target area.

[0052] The specific technical solutions adopted by the present invention include:

[0053] (1) Fitting of overall probability distribution of typhoon parameters

[0054] The present invention uses the kernel density estimation method to fit the historical sample set of tropical cyclones in the northwest Pacific Ocean (including the South China Sea, north of the equator, and west of 180° east longitude) based on the 1949-2018 typhoon best path data and 1980-2016 typhoon scale data provided by the Tropical Cyclone Data Center of the China Meteorological Administration. The sample set contains the typhoon location information recorded every 6 hours, the maximum wind speed near the typhoon center (2-minute average), and the typhoon intensity V calculated from it. m , Coriolis coefficient f, typhoon moving speed V t and the typhoon moving direction angle θ. Among them, the scale information corresponds to the 34 nautical miles / hour wind circle radius, that is, V * =The radius r corresponding to 17.49m / s * value.

[0055] For a sample set X = [x1, x2, ..., x i ,…,x n ], and its probability density value at x is calculated by formula 1a:

[0056]

[0057] h is the neighborhood range parameter of x, N(x i ∈[x–h,x+h]) represents the number of sample points in the neighborhood, and n is the total number of samples. Formula 1a is further expressed as Formula 1b:

[0058]

[0059] in, K(x) represents the kernel function; 1(x≤1) means that it takes the value of 1 when x≤1, otherwise it takes the value of 0; integrating the probability density function shown in formula 1b, we get formula 1c:

[0060]

[0061] It can be seen from the above formula that as long as the integral of the kernel function K is equal to 1, the characteristic of the probability distribution function being equal to 1 can be satisfied. Therefore, in practical applications, a known function form is selected as the kernel function K, and the function form satisfies the condition that the integral is equal to 1; in this application, the standard normal distribution is selected, and the probability density value is further expressed as formula 1d:

[0062]

[0063] φ is the density function of the standard normal distribution. The above formula shows that the probability density at x is expressed as: n normal distributions (the distribution means are x i , standard deviation is h) is the weighted sum of the function values ​​at x. For its specific mathematical meaning, see Figure 1 .

[0064] The present invention adopts Gaussian kernel function as the kernel function form. h is also commonly referred to as bandwidth. For Gaussian kernel function, the h value represents the standard deviation of the kernel function. Usually, the h value is adjusted according to the fitting situation in combination with the histogram characteristics. In the following typhoon key parameter analysis, the generalized extreme value distribution and non-parametric kernel density estimation method are used to fit the probability distribution of typhoon parameters. Figure 2 The maximum wind speed of the typhoon is given in turn. m , Coriolis coefficient f, typhoon moving speed V t The distribution and fitting of the typhoon moving direction angle θ. Compared with the generalized extreme value distribution, the nonparametric kernel density estimation method can more accurately describe the distribution characteristics of historical samples.

[0065] According to the typhoon scale dataset, Figure 3 The typhoon related scale r is given * (The corresponding V * =17.49m / s) and the distribution and fitting of the typhoon absolute scale r0. Figure 3 It can be seen that the typhoon-related scale r * The statistical distribution of and absolute scale r0 can be effectively fitted by generalized extreme value distribution; r0 is an independent free variable, and its probability characteristics can be described by lognormal distribution. Figure 3 As can be seen from Figure 2, in the northwest Pacific region, the distribution of the absolute scale of typhoons r0 can also be effectively fitted by the lognormal distribution. The fitting results of the generalized extreme value distribution, lognormal distribution and kernel density estimation are very close.

[0066] (2) Sampling of typhoon starting positions

[0067] Based on the annual frequency of typhoons, the typhoon starting position is sampled. Based on the typhoon best path dataset, the following Figure 4 The annual frequency distribution is shown in Figure 2. The average annual number of typhoons in the northwest Pacific region is about 35. Figure 4 Poisson distribution and negative binomial distribution were used to fit the annual frequency. From the fitting results, the variance of Poisson distribution fitting is smaller than that of negative binomial distribution. This shows that if Poisson distribution is used, it is more likely to draw the average annual frequency and its nearby values ​​when sampling the annual frequency; while if negative binomial distribution is used, it is more likely to draw the annual frequency values ​​at both ends of the distribution when sampling the annual frequency.

[0068] For the probability distribution fitting of the typhoon starting position, the present invention adopts a two-dimensional nonparametric Gaussian kernel density estimation method, and the fitting results are as follows: Figure 4 As shown. Figure 4 b It can be seen that, with the two main typhoon occurrence centers in the northwest Pacific region (16.5°N, 117°E) and (9°N, 142°E) as the center, the probability of typhoon occurrence gradually decreases as the range increases. Figure 4 The location probability distribution shown in b, combined with the sampling results of the annual occurrence frequency, can be used to obtain the starting location information of typhoons each year through random sampling, thus completing the typhoon starting location sampling.

[0069] (3) Typhoon movement prediction based on Markov chain

[0070] The Markov chain method is used to simulate the typhoon's moving process, and the non-parametric kernel density estimation method is used to fit the probability distribution of the typhoon's parameters during the moving process to ensure the continuity of the typhoon's path. The memoryless characteristic of the Markov chain in the time scale is expressed by formula 2a:

[0071] P(X n+m =j|X n =i n ,X n-1 =i n-1 ,K X1=i1)=P(X n+m =j|X n =i n )(2a)

[0072] X n represents a Markov chain, and m represents the time interval. When m = 1, the above formula holds. nThe countable set of all possible values ​​of is called the state space of the chain, denoted by S. By mathematical induction, it can be proved that when m is any positive integer, formula 2a still holds. The second property of the Markov chain is called time homogeneity (the probability of the system changing from state i to state j depends only on the length of the time interval), as shown in formula 2b:

[0073] P(X n+m =j|X n =i) = P(X n+m+k =j|X n+k =i)(2b)

[0074] Since the conditional probability P(X n+m =j|X n =i) is independent of the starting step n and is further written as p ji (m). Probability p ji (m) represents the transition probability from state i to state j. ji (m) is an element, and the m-step state transition probability matrix P(m) of the Markov chain is constructed, as shown in Formula 2c:

[0075]

[0076] The elements in the matrix P are all non-negative, and the sum of the elements in each column is equal to 1. Among them, i and j represent the system state and belong to the state space S. Take m = 1, then the probability distribution of the system state in the n+1 time step is calculated by formula 2d:

[0077] Y n+1 =PY n (2d)

[0078] According to the memoryless property of Markov chain, formula 2d is further expressed as formula 2e:

[0079] Y n+1 =P n+1 Y 0 (2e)

[0080] Among them, Y n represents the probability vector of the system state at n time steps; Y 0 It represents the probability distribution of the initial state of the system. Based on the above description, we can see that the Markov chain is represented by a triple, namely (S, P, Y 0 ). S is the state space, that is, the set of all possible states; P is the state transfer matrix; Y 0 is the probability distribution of the initial state. Given these three tuples, we can construct the corresponding Markov chain.

[0081] The Monte Carlo method based on Markov chain is used to simulate the typhoon's movement. When constructing the state transfer matrix, the rate of change of typhoon parameters is used to describe the state transition. The typhoon intensity (i.e. the maximum wind speed V m ) as an example, its state transition probability is expressed as To represent. Among them, V m is the typhoon intensity value at the current moment; is the rate of change of the typhoon intensity in the next time step relative to the current step. Then, The current typhoon intensity is V m When the intensity changes at the next time step, Since the parameter of rate of change is introduced, a new tuple needs to be added when generating the Markov chain, namely the state change rate space R. Then, the Markov chain in the typhoon simulation is represented by a four-tuple, namely (S, R, P, Y 0 ). Based on the analysis results of the typhoon initial simulation, the initial state probability distribution Y of each typhoon parameter is obtained. 0 The state space S and the state change rate space R need to be obtained based on the statistics of historical typhoon data. Based on the historical typhoon path data, the non-parametric kernel density estimation method is then used to create the state transition matrix. As mentioned earlier, the state transition probability belongs to conditional probability. Based on the property of conditional probability, taking typhoon intensity as an example, its state transition probability can be further expressed as formula 2f.

[0082]

[0083] If the typhoon intensity and its rate of change are regarded as continuous variables, then the state transition probability Equal to the joint probability With marginal probability P(V m ). Based on historical data, this application uses a two-dimensional nonparametric kernel density method to fit the joint density of typhoon parameters and their change rates, and uses a one-dimensional nonparametric kernel density estimation method to fit the corresponding marginal probabilities. Here, the historical data of typhoon travel are divided into two groups: ocean data and land data, and statistical analysis and probability modeling are performed on these two groups of data respectively. Figure 6 to Figure 9 The marginal distribution of each typhoon parameter and its joint distribution with its changing rate are given.

[0084] (4) Determining whether a typhoon has ended

[0085] According to the typhoon location information, a typhoon termination probability model is constructed. In the full typhoon simulation, the typhoon termination judgment is connected with the typhoon advance simulation. Every time the typhoon advances, it is necessary to judge whether it terminates at the current step. The typhoon termination probability model can be expressed as formula 3a.

[0086] P(E=0|x,y)+P(E=1|x,y)=1(3a)

[0087] Among them, x represents the latitude of the typhoon center; y represents the longitude of the typhoon center; E=0 means the typhoon has terminated; E=1 means the typhoon continues to move. Here, the zoning statistics method is used to construct a spatially discrete typhoon termination probability model. The specific operations are as follows: ① Define the termination area range (latitude 0°~60°N, longitude 100°E~180°E); ② Divide the grid within the termination area, where the grid density is set to 2°×2°; ③ Divide the historical termination point samples into two categories: sea surface termination points and land termination points, and perform probability modeling separately; ④ Take all historical typhoon path points as the object, and count the number of path points in each grid; ⑤ Take the historical termination points as the object, and count the number of termination points in each grid; ⑥ The ratio of the number of termination points to the number of path points is the typhoon termination probability of the corresponding grid; ⑦ If there is no historical path point in the grid, its termination probability is set to 1; if there is a historical path point but no historical termination point, the termination probability is 0. According to the above method, Fig.10 The spatial distribution of the termination probability values ​​of sea typhoons and land typhoons is given.

[0088] (5) Monte Carlo simulation of the entire typhoon path

[0089] The Monte Carlo simulation of the entire typhoon path involves three steps: starting simulation, moving simulation and ending simulation, which includes five state information: typhoon position, typhoon intensity, moving speed, moving direction and typhoon scale. Fig.11 The process shown is implemented. Among them, i represents the year number; j represents the typhoon number that occurred in a year; k represents the time number of a typhoon during its movement; Δt represents the time interval between adjacent time numbers, which is 6 hours; (x, y) represents the latitude and longitude of the typhoon center; the reckon function represents the calculation of the typhoon position at the next moment based on the current position, moving speed and moving direction angle information.

[0090] (6) Extraction of typhoon parameter information in the target area

[0091] Combining the random simulation results of the typhoon process and the typhoon wind field simulation method, the typhoon parameter information of the target area is calculated through the following steps:

[0092] 1. Screening of typhoon path points:

[0093] Based on the 10,000-year typhoon simulation results in the target area, find the typhoon path points whose distance from the typhoon center to the target location does not exceed the corresponding typhoon absolute scale and their corresponding typhoon sequence number and occurrence year sequence number. The found path points, typhoon sequence number and occurrence year number are the objects of the next examination;

[0094] Count the typhoon serial numbers j (j = 1, 2, …, J) that affect the target area in the i-th year (i = 1, 2, …, I); obtain the typhoon path points k (k = 1, 2, …, K) that affect the target area in the j-th typhoon;

[0095] 2. Typhoon wind field calculation:

[0096] According to the typhoon state parameters (including center position, intensity, scale, moving speed and moving direction) of the path point k, use the typhoon wind field numerical simulation method to calculate the surface wind field (i.e., wind speed and wind direction) of the target area and record it; if k < K, let k = k + 1 and return to the previous step; if k = K, continue to the next step;

[0097] 3. Typhoon wind speed extreme value extraction:

[0098] Count the wind field data caused by all path points k (k = 1, 2, …, K) in the j-th typhoon on the target area, find the maximum wind speed and its corresponding wind direction, and record them as the wind speed extreme value and the corresponding wind direction caused by the j-th typhoon; if j < J, let j = j + 1 and return to obtaining the typhoon path points k (k = 1, 2, …, K) that affect the target area in the j-th typhoon;

[0099] Based on the above steps, the typhoon parameter information of the target area can be finally obtained.

[0100] (7) Typhoon wind field calculation based on the Emanuel physical model

[0101] The surface wind is divided into two components: one is the axisymmetric vortex wind related to the typhoon vortex The other is the surface background wind related to the local environment Surface wind Is expressed as the weighted sum of the two components, as shown in

[0102] Equation 4a.

[0103]

[0104] Where α r Is the reduction coefficient of the vortex wind, and α b Is the participation coefficient of the background wind. Among them, the vortex wind The magnitude of which is calculated by the Emanuel typhoon radial profile model, and the inflow angle α is introduced to correct its direction. The inflow angle α is calculated by equations 4b, 4c and 4d:

[0105]

[0106]

[0107] α=25o,r≥1.2r m (4d)

[0108] Background Wind The magnitude is equal to the typhoon's moving speed V t , and introduce a rotation angle β to correct its direction. For northern hemisphere typhoons, β is taken as 20° counterclockwise. Based on the above explanation, the surface wind caused by northern hemisphere typhoons can be Fig.12 In the figure, T represents the target position. Fig.12 The surface wind conversion shown can obtain the surface typhoon wind field information.

[0109] (8) Conversion of near-surface typhoon wind field

[0110] In the right Fig.12 When calculating the surface wind at the target position T, the distance r needs to be judged first: when r ≥ r a , using the typhoon external radial profile model to calculate the typhoon vortex wind speed V r ; When r <r a , the typhoon internal radial profile model is used to calculate the typhoon vortex wind speed V r The vortex wind speed solved by the typhoon radial model is the average wind speed within the thickness of the boundary layer. However, the surface wind speed usually required is the wind speed at a height of 10m from the sea surface or the ground surface. Therefore, when converting the vortex wind into the surface wind, a reduction factor α is introduced. r , the value is 0.85. Background wind V t The participation coefficient α b The value is 0.55. Due to the lack of relevant statistical research on land typhoons, the α value of offshore typhoons is still used after the typhoon lands. r Experience value 0.85 and α b Experience value: 0.55.

[0111] As the typhoon moves from the sea to the ground, the environmental parameters involved in the typhoon model (surface drag coefficient and the ratio of the surface heat entropy exchange coefficient to the drag coefficient) change. Here, the method of updating environmental parameters is used to distinguish the calculation of offshore typhoons and land typhoons. Before and after the typhoon lands, the surface drag coefficient C D The calculation method of the ratio of the surface heat entropy exchange coefficient to the resistance coefficient ξ is shown in Table 1:

[0112] Table 1 Environmental parameters of the radial profile of Typhoon Emanuel

[0113]

[0114] When the typhoon center is located on the sea surface, the ocean surface wind field obtained by calculating according to Formula 4a needs to be further converted into the land surface wind field. Here, a reduction coefficient of 0.85 is introduced to convert the calculated typhoon wind speed on the sea surface into the typhoon wind speed on the ground surface.

[0115] In addition, the typhoon meteorological data in China currently gives the 2-minute average measurement value. In practical applications, the 10-minute average wind speed is usually adopted. This requires time interval conversion for the surface wind speed obtained by calculating according to Formula 4a. The wind speed time interval conversion relationship here is as shown in Formula 5a.

[0116]

[0117] The subscript 3600 on the right side of Formula 5a represents the time interval, with the unit of seconds; z is the height, with the unit of meters; z0 is the surface roughness length; ε is an empirical coefficient related to the roughness length; c(t) is a statistical parameter related to the time interval. When t = 2 min, c = 0.956, and when t = 10 min, c = 0.36

[128] . Then, the conversion relationship between the 10-minute average wind speed and the 2-minute average wind speed at a height of 10 m can be further expressed as Formula 5b.

[0118]

[0119] Among them, ε is determined according to z0.

[0120] (9) Generation of annual wind speed extreme value and its wind direction data set

[0121] 1) According to the typhoon simulation results of 10,000 years, find the typhoon path points where the distance from the typhoon center point to the target location does not exceed the corresponding typhoon absolute scale, as well as their corresponding typhoon serial numbers and occurrence year serial numbers. The found path points, typhoon serial numbers, and occurrence year numbers are the objects to be examined next;

[0122] 2) Count the typhoon serial numbers j (j = 1, 2,..., J) that affect the target location in the i-th year (i = 1, 2,..., I);

[0123] 3) Obtain the typhoon path points k (k = 1, 2,..., K) that affect the target location in the j-th typhoon;

[0124] 4) According to the typhoon state parameters (including the center position, intensity, scale, moving speed, and moving direction) of the path point k, use the typhoon wind field numerical simulation method to calculate the surface wind field (i.e., wind speed and wind direction) of the target area and record it;

[0125] 5) If k < K, let k = k + 1 and return to step (4); if k = K, continue to the next step;

[0126] 6) Statistically analyze the wind field data caused by all path points k (k = 1, 2, …, K) in the j-th typhoon to the target area, find the maximum wind speed and its corresponding wind direction, and record them as the wind speed extreme value and the corresponding wind direction caused by the j-th typhoon;

[0127] 7) If j < J, let j = j + 1 and return to step (3); if j = J, continue to the next step;

[0128] 8) Statistically analyze the wind speed extreme values and their wind directions caused by all typhoons j (j = 1, 2, …, J) in the i-th year, find the maximum wind speed extreme value and its corresponding wind direction, and record them as the annual wind speed extreme value and the corresponding wind direction caused by the typhoons in the i-th year;

[0129] 9) If i < I, let i = i + 1 and return to step (2); if i = I, continue to the next step;

[0130] 10) Statistically analyze the annual wind speed extreme values and their wind directions of all year numbers i (i = 1, 2, …, I) and record them for constructing the typhoon wind speed and typhoon wind field probability model in the target area.

[0131] Combined with the stochastic simulation results of the typhoon process and the typhoon wind field simulation method, obtain the typhoon annual speed extreme value samples and their corresponding wind direction samples in the example area (Zhanjiang City). Fig.13 The distribution of the typhoon annual wind speed extreme value and the corresponding wind direction in this area is given. Fig.13 a. The wind speed distribution was respectively fitted using the generalized extreme value distribution and the kernel density estimation method. After the fitting results and probability vulnerability assessment, the fitting results of the kernel density estimation method are adopted here. From Fig.13 b. From the given wind direction probability distribution, the typhoon wind directions in the example area take values within the range of 0° to 360°. Therefore, the wind direction is regarded as a continuous random variable here, and the kernel density estimation method is used to fit its distribution.

[0132] Based on the above statistical situation of the annual extreme values of the typhoon wind field, the probability distribution characteristics of the wind speed and wind direction can be further represented by Fig.14 shown. Fig.14 a. The cumulative probability distribution of the wind speed is given, where P represents the cumulative probability value. Assume that the probability that the wind speed extreme value caused by a typhoon in a year is less than the wind speed extreme value corresponding to the return period of R years is 1 - 1 / R. Then, based on the cumulative probability distribution in 14a, the typhoon wind speed extreme value corresponding to the return period of R years can be calculated. Fig.14The maximum wind speed of typhoons with a return period of 10 years, 50 years, 100 years and 500 years is shown in a, and its values ​​are 27.72m / s, 38.23m / s, 42.43m / s and 52.36m / s respectively. According to the wind speed and wind pressure conversion method provided by the building wind load specification, the basic wind pressure values ​​corresponding to the return periods of 10 years, 50 years, 100 years and 500 years in the example area can be obtained, which are: 0.48kN / m 2 、0.91kN / m 2 , 1.12kN / m 2 , 1.71kN / m 2 According to my country's building load code, the basic wind pressure corresponding to the return period of 10, 50 and 100 years in Zhanjiang City is: 0.5kN / m 2 、0.8kN / m 2 and 0.95kN / m 2 The simulation results were compared with the historical statistical results in the code, and it was found that the simulated wind speed extreme value of 10-year return period was slightly smaller than the historical statistical results; the simulated wind speed extreme values ​​of 50-year and 100-year return periods were slightly larger than the historical statistical results. Fig.14 b The probability distribution of wind direction corresponding to the annual wind speed extreme value is further expressed in polar coordinates. Through polar coordinates, it is found that the typhoon wind direction in Zhanjiang area is mainly northerly, especially in the three directions of north northwest (NNW), due north (N) and north northeast (NNE).

[0133] Fig.13 and Fig.14 In both cases, the wind speed extreme value and wind direction are considered separately, and then their respective probability distributions are obtained. In the typhoon probability vulnerability assessment of transmission lines, the wind speed extreme value and wind direction are a set of independent random variables that jointly affect the vulnerability of transmission towers in the line. Therefore, based on the typhoon numerical simulation results, the two-dimensional Gaussian nonparametric kernel density estimation method is used to construct a joint probability model of annual wind speed extreme value and wind direction. Fig.15 The joint probability density of wind speed and wind direction is given.

[0134] The complete steps of the method of the present invention are as follows:

[0135] (I) Numerical simulation of typhoon wind field

[0136] 1) Establish a typhoon radial profile model

[0137] The Emanuel typhoon radial profile model is adopted, assuming that the air angular momentum and the partial derivative of the angular momentum with respect to the radius are continuous at the inner and outer junctions of the typhoon radial profile. By introducing the continuity condition of the air angular momentum at the inner and outer junctions, the full radial profile model of the typhoon can be finally obtained. It includes: 3 environmental parameters of the Emanuel typhoon radial profile model (i.e., sea or surface resistance coefficient CD , the sinking rate w of the free tropospheric air in the typhoon outer area and the ratio of the surface heat entropy exchange coefficient to the surface drag coefficient ξ) and six typhoon parameters (i.e., the maximum wind speed V m and its corresponding radius r m , the radius of the intersection of the inner and outer regions is r a And the corresponding wind speed V a , typhoon scale r0 and latitude of typhoon center ). Then introduce the typhoon peripheral observation point (r * ,V * ), the full radial profile of the typhoon can be obtained by using the method of internal and external joint solution. The steps are: (1) solve the external radial profile (2) solve the internal and external intersection point (r a ,V a ) and the internal radial section (3) are connected with the internal and external radial sections according to the external section, internal section and intersection value obtained in steps (1) and (2).

[0138] 2) Calculate the near-surface typhoon wind speed

[0139] Considering that typhoon vortices are often asymmetric, the surface wind is divided into two components: one is the axisymmetric vortex wind V related to the typhoon vortex. r ; The second is the surface background wind V related to the local environment t The surface wind U is expressed as the weighted sum of the two components. The size of the vortex wind Vr is calculated by the radial profile model of Typhoon Emanuel, and the inflow angle α is introduced to correct its direction. Background wind V t The magnitude is equal to the typhoon's moving speed V t , and introduce the rotation angle β to correct its direction. In most cases, the transmission tower-line system is located on the land surface rather than the sea surface, so the sea surface typhoon wind speed is converted to the surface typhoon wind speed, and the obtained surface wind speed is converted into a time-distance conversion.

[0140] 3) Average wind speed profile

[0141] The vertical profile of the average wind speed is described by a logarithmic law function. Taking into account the influence of the boundary layer thickness on the applicable range of the logarithmic law, the empirical formula of the vertical profile is modified. According to the difference in average wind speed at different times and heights. The average wind speed U at time t at height z is equal to the product of the vertical profile of the average wind speed and the time function f(t). The time function f(t) is called the time profile of the average wind speed. Set U(z) as the maximum average wind speed value at height z during the typhoon process, then the time profile f(t) is a function less than or equal to 1. The specific expression of the time profile f(t) is affected by the spatiotemporal characteristics of the typhoon throughout its entire process.

[0142] 4) Constructing a pulsating wind model

[0143] The fluctuating wind of a typhoon is usually regarded as a non-stationary random process. It is assumed to be a Gaussian non-stationary process, and the evolving power spectral density method (EPSD) is used to simulate the fluctuating wind. It includes: assuming that the frequency composition of the fluctuating wind remains unchanged throughout the typhoon, calculating the time-frequency conversion relationship and the evolving power spectral density function of the non-stationary fluctuating wind speed, and then expressing the evolving power spectral density of the non-stationary fluctuating wind by the modulation function and the power spectral density of a certain stationary process. Then, the correlation of the fluctuating wind at two points in space along the vertical z and horizontal x is considered, and the coherence function proposed by Davenport is used. Finally, combined with the structural space examination point, the evolving power spectral density matrix of the non-stationary fluctuating wind speed of the typhoon can be obtained. Based on the above fluctuating wind spectrum, the harmonic synthesis method based on the superposition of trigonometric series is used to simulate the non-stationary process of the fluctuating wind. Here, the Deodatis harmonic synthesis method is used to simulate the fluctuating wind field of the transmission tower-line system. After calculation, the Fourier transform technology is used to solve it. Finally, the time history of non-stationary fluctuating wind can be obtained; combined with the above content, the evolution power spectrum of non-stationary fluctuating wind can be obtained.

[0144] (II) Stochastic simulation of typhoon process

[0145] 1) Calculate the overall probability distribution of typhoon parameters

[0146] There are five key parameters involved in solving the radial wind speed profile. They are: typhoon intensity V m 、Typhoon related scale r * (or absolute scale r0), Coriolis coefficient f, typhoon moving speed V t and the typhoon moving direction angle θ. Then, according to the typhoon optimal path data and typhoon scale data, the typhoon intensity V is obtained. m , Coriolis coefficient f, typhoon moving speed V t and historical statistical data of the typhoon moving direction angle θ. For the historical sample set, a non-parametric estimation method is used to fit the distribution of the sample set. Kernel density estimation is a type of non-parametric estimation. The Gaussian kernel function is used to estimate the maximum wind speed V of the typhoon. m , Coriolis coefficient f, typhoon moving speed V t And the typhoon moving direction angle θ is distributed and fitted.

[0147] 2) Adopt the whole process random simulation method

[0148] The random simulation method of the whole process of typhoon is divided into: initial simulation, progress simulation and termination simulation.

[0149] The typhoon start simulation can be summarized into three steps: ① annual frequency sampling; ② based on the annual frequency, extract the starting position information of each typhoon; ③ extract the typhoon's starting intensity, starting moving speed, starting moving direction angle and starting scale information in turn. The essence of typhoon start simulation is a random sampling simulation based on the starting point probability model. The probability model involved can be summarized in Table 2.

[0150] Table 2 Probability model for typhoon onset simulation

[0151]

[0152] The typhoon movement simulation can be achieved through the following four steps: ① According to the initial state probability model, the initial values ​​of the typhoon parameters are randomly extracted; ② According to the state information of the typhoon at the current time step, combined with the state transition probability model (joint probability density and marginal density) corresponding to each parameter, the state change rate of the typhoon parameters for the next time step is randomly extracted and calculated; ③ According to the state change rate sampling results, the typhoon state parameters and the typhoon center position of the next time step are obtained; ④ The state information of the next time step is saved as the current state of the typhoon, and then return to step ② to perform the next step of typhoon state simulation.

[0153] Termination simulation uses the zoning statistics method to construct a spatially discrete typhoon termination probability model. The specific operations are as follows: ① Delineate the scope of the termination area; ② Divide the grid within the termination area; ③ Divide the historical termination point samples into two categories: sea termination points and land termination points, and perform probability modeling respectively; ④ Take all historical typhoon path points as the object, and count the number of path points in each grid; ⑤ Take the historical termination points as the object, and count the number of termination points in each grid; ⑥ The ratio of the number of termination points to the number of path points is the typhoon termination probability of the corresponding grid; ⑦ If there is no historical path point in the grid, its termination probability is set to 1; if there is a historical path point but no historical termination point, the termination probability is 0.

[0154] (III) Probability distribution of typhoon wind field

[0155] The typhoon wind speed extreme value and its wind direction can be calculated by the following steps:

[0156] (1) Based on the 10,000-year typhoon simulation results in the calculation area, find the typhoon path points whose distance from the typhoon center to the calculation area does not exceed the corresponding typhoon absolute scale and their corresponding typhoon sequence number and occurrence year sequence number. The found path points, typhoon sequence number and occurrence year number are the objects of the next examination;

[0157] (2) Count the typhoon number j (j = 1, 2, ..., J) that affects the calculation area in year i (i = 1, 2, ..., I);

[0158] (3) Obtain the typhoon path points k (k = 1, 2, …, K) in the j-th typhoon that affect the calculation area;

[0159] (4) According to the typhoon state parameters of path point k (including the central position, intensity, scale, moving speed, and moving direction), use the numerical simulation method of typhoon wind field to calculate the surface wind field (i.e., wind speed and wind direction), and record it;

[0160] (5) If k < K, let k = k + 1 and return to step (4); if k = K, continue to the next step;

[0161] (6) Statistically analyze the wind field data caused by all path points k (k = 1, 2, …, K) in the j-th typhoon on the calculation area, find the maximum wind speed and its corresponding wind direction, and record them as the wind speed extreme value and the corresponding wind direction caused by the j-th typhoon;

[0162] (7) If j < J, let j = j + 1 and return to step (3); if j = J, continue to the next step;

[0163] (8) Statistically analyze the wind speed extreme values and their wind directions caused by all typhoons j (j = 1, 2, …, J) in the i-th year, find the maximum wind speed extreme value and its corresponding wind direction, and record them as the annual wind speed extreme value and the corresponding wind direction caused by the typhoons in the i-th year;

[0164] (9) If i < I, let i = i + 1 and return to step (2); if i = I, continue to the next step;

[0165] (10) Statistically analyze the annual wind speed extreme values and their wind directions of all year numbers i (i = 1, 2, …, I) and record them for constructing the typhoon wind field probability model.

[0166] Based on the above steps, the typhoon annual wind speed extreme value samples for 10,000 years and their corresponding wind direction samples can be finally obtained.

[0167] Demonstration of the utility of the method of the present invention:

[0168] Based on the typhoon wind field simulation and the typhoon process stochastic simulation (Monte Carlo Markov chain method), taking the typhoons suffered by Zhanjiang area in 10,000 years as an example, the typhoon annual wind speed extreme value samples for 10,000 years and their corresponding wind direction samples in Zhanjiang area can be finally obtained. Figure 4 The distribution of the typhoon annual wind speed extreme value and the corresponding wind direction in Zhanjiang area is given. From Figure 4 From the given wind direction probability distribution, the typhoon wind directions in Zhanjiang area take values in the range of 0° to 360°. Therefore, the wind direction is regarded as a continuous random variable here, and the kernel density estimation method is used to fit its distribution.

[0169] Based on the above statistics of the annual extreme values ​​of typhoon wind fields, the probability distribution characteristics of wind speed and wind direction can be further analyzed by Figure 5 express. Figure 5 a gives the cumulative probability distribution of wind speed, where P represents the cumulative probability value. Assuming that the probability that the extreme wind speed caused by a typhoon in a year is less than the corresponding extreme wind speed with a return period of R years is 1-1 / R. Then, based on the cumulative probability distribution of 5a, the extreme wind speed of the typhoon corresponding to the return period of R years can be calculated. Figure 5 The maximum wind speed of typhoons with a return period of 10 years, 50 years, 100 years and 500 years is shown in a, and its values ​​are 27.72m / s, 38.23m / s, 42.43m / s and 52.36m / s respectively. According to the wind speed and wind pressure conversion method provided by the building wind load specification, the basic wind pressure values ​​corresponding to the return periods of 10 years, 50 years, 100 years and 500 years in Zhanjiang area can be obtained, which are: 0.48kN / m 2 、0.91kN / m 2 , 1.12kN / m 2 , 1.71kN / m 2 According to my country's building load code, the basic wind pressure corresponding to the return period of 10, 50 and 100 years in Zhanjiang City is: 0.5kN / m 2 、0.8kN / m 2 and 0.95kN / m 2 Comparing the simulation results with the historical statistical results in the code, it is found that the simulated wind speed extreme value of 10-year return period is slightly smaller than the historical statistical results; however, the simulated wind speed extreme values ​​of 50-year and 100-year return periods are slightly larger than the historical statistical results. Figure 5 b The probability distribution of wind direction corresponding to the annual wind speed extreme value is further expressed in polar coordinates. Through polar coordinates, it is found that the typhoon wind direction in Zhanjiang area is mainly northerly, especially in the three directions of north northwest (NNW), due north (N) and north northeast (NNE).

[0170] Considering that in the typhoon probabilistic vulnerability assessment of transmission lines, the wind speed extreme value and wind direction are a set of independent random variables that jointly affect the vulnerability of transmission towers in the line. Therefore, based on the typhoon numerical simulation results, the two-dimensional Gaussian nonparametric kernel density estimation method is used to construct a joint probability model of annual wind speed extreme value and wind direction. Figure 6 The joint probability density of wind speed and wind direction is given.

[0171] The method of the present invention can realize the joint probability distribution of wind speed and wind direction. The probability model can be applied to subsequent probabilistic vulnerability assessment and has significant effectiveness and applicability.

[0172] Finally, it should be noted that the above description is only a preferred example of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for generating the joint probability distribution of wind speed and wind direction of a strong typhoon based on Monte Carlo Markov chain, characterized by: The following steps are involved: S1. First, the typhoon radial profile model of Emanuel typhoon and the near-surface typhoon wind speed conversion method, as well as the average wind vertical profile model and the pulsating wind model are established to achieve the numerical simulation of the typhoon wind field; S2. Then, the typhoon process is stochastically simulated using the method of stochastic simulation of the entire process, including the initial simulation, progress simulation and termination simulation, to conduct stochastic simulation of the typhoon process and analyze the simulation results, and the typhoon position at the next time step is calculated by combining the simulation methods of each stage; S3. Finally, the two-dimensional Gaussian nonparametric kernel density estimation method is used to calculate the probability distribution of the typhoon wind field to obtain the annual wind speed extreme value and wind direction data set, and to construct a joint probability model of strong typhoon wind speed and wind direction.

2. The method for generating the joint probability distribution of wind speed and wind direction of a strong typhoon based on a Monte Carlo Markov chain according to claim 1, characterized in that: In step S1, establishing the radial profile model of Typhoon Emanuel specifically includes: Assuming that the air angular momentum and the partial derivative of angular momentum with respect to radius are continuous at the inner and outer junctions of the typhoon radial section, by introducing the continuity condition of air angular momentum at the inner and outer junctions, the typhoon full radial section model is finally obtained. The typhoon full radial section model includes three environmental parameters of the Emanuel typhoon radial section model, namely, the sea or surface resistance coefficient C D , the sinking rate w of the free tropospheric air in the outer area of ​​the typhoon and the ratio of the surface heat entropy exchange coefficient to the surface drag coefficient ξ, as well as 6 typhoon parameters, namely the maximum wind speed V of the typhoon m and its corresponding radius r m , the radius of the intersection of the inner and outer regions is r a And the corresponding wind speed V a , typhoon scale r0 and latitude of typhoon center Then introduce the typhoon outer observation point (r * ,V * ), the internal and external joint solution method is used to obtain the full radial profile of the typhoon.

3. The method for generating the joint probability distribution of wind speed and wind direction of a strong typhoon based on a Monte Carlo Markov chain according to claim 1, characterized in that: In step S1, the near-surface typhoon wind speed is calculated, and the surface wind is divided into two components: one is the axisymmetric vortex wind V related to the typhoon vortex r ; The second is the surface background wind V related to the local environment t ; The surface wind U is expressed as the weighted sum of the two components; the size of the vortex wind Vr is calculated by the Emanuel typhoon radial profile model, and the inflow angle α is introduced to correct its direction; the background wind V t The magnitude is equal to the typhoon's moving speed V t , and introduce a rotation angle β to correct its direction.

4. The method for generating the joint probability distribution of wind speed and wind direction of a strong typhoon based on a Monte Carlo Markov chain according to claim 1, characterized in that: In step S1, the vertical profile of the average wind speed is described by a logarithmic function, and the empirical formula of the vertical profile is corrected; according to the difference in average wind speed at different times and heights; the average wind speed U at time t at height z is equal to the product of the vertical profile of the average wind speed and the time function f(t); the time function f(t) is called the time profile of the average wind speed; U(z) is set to the maximum average wind speed value at height z during the typhoon history, then the time profile f(t) is a function less than or equal to 1.

5. The method for generating the joint probability distribution of wind speed and wind direction of a strong typhoon based on a Monte Carlo Markov chain according to claim 1, characterized in that: In step S1, the steps of constructing a fluctuating wind model include: The fluctuating wind of a typhoon is assumed to be a Gaussian non-stationary process, and the evolutionary power spectrum density method is used to simulate the fluctuating wind, including assuming that the frequency composition of the fluctuating wind remains unchanged throughout the typhoon process, calculating the time-frequency conversion relationship and the evolutionary power spectrum density function of the non-stationary fluctuating wind speed, and then expressing the evolutionary power spectrum density of the non-stationary fluctuating wind by the modulation function and the power spectrum density of a certain stationary process. Then, the correlation of the fluctuating wind at two points in space along the vertical z and horizontal x is considered, and the coherence function proposed by Davenport is used. Finally, the evolutionary power spectrum density matrix of the non-stationary fluctuating wind speed of the typhoon is obtained in combination with the structural space examination points. Based on the above fluctuating wind spectrum, the harmonic synthesis method based on the superposition of trigonometric series is used to simulate the non-stationary process of the fluctuating wind. The Deodatis harmonic synthesis method is used to simulate the fluctuating wind field of the transmission tower-line system. After calculation, the Fourier transform technology is used to solve it. Finally, the evolutionary power spectrum of the non-stationary fluctuating wind time course is obtained.

6. The method for generating the joint probability distribution of wind speed and wind direction of a strong typhoon based on a Monte Carlo Markov chain according to claim 1, characterized in that: In step S2, the initial simulation steps of the typhoon include: ① annual occurrence frequency sampling; ② based on the annual occurrence frequency, extracting the starting position information of each typhoon; ③ extracting the typhoon's initial intensity, initial moving speed, initial moving direction angle and initial scale information in sequence.

7. The method for generating the joint probability distribution of wind speed and wind direction of a strong typhoon based on a Monte Carlo Markov chain according to claim 1, characterized in that: In step S2, the typhoon movement simulation step includes: ① randomly extracting the initial values of each typhoon parameter according to the initial state probability model; ② randomly extracting and calculating the state change rate of the typhoon parameter for the next time step according to the state information of the typhoon at the current time step and combining the state transition probability model corresponding to each parameter; ③ obtaining the typhoon state parameter and the typhoon center position for the next time step according to the sampling result of the state change rate; ④ saving the state information of the next time step as the current state of the typhoon, and then returning to step ② for the next typhoon state simulation.

8. The method for generating the joint probability distribution of wind speed and wind direction of a strong typhoon based on a Monte Carlo Markov chain according to claim 1, characterized in that: In step S2, the termination simulation adopts the zonal statistics method to construct a spatially discrete typhoon termination probability model. The specific steps include: ① defining the range of the termination area; ② dividing grids within the range of the termination area; ③ classifying the historical termination point samples into two categories: sea surface termination points and land termination points, and performing probability modeling separately; ④ taking all historical typhoon path points as objects, and counting the number of path points in each grid; ⑤ taking the historical termination points as objects, and counting the number of termination points in each grid; ⑥ the ratio of the number of termination points to the number of path points is the typhoon termination probability of the corresponding grid; ⑦ if there are no historical path points in the grid, its termination probability is set to 1; if there are historical path points but no historical termination points, the termination probability is 0.

9. The method for generating the joint probability distribution of wind speed and wind direction of a strong typhoon based on a Monte Carlo Markov chain according to claim 1, characterized in that: In step S3, the annual extreme wind speed and wind direction data set is calculated through the following steps: 1) According to the typhoon simulation results of 10,000 years, find the typhoon path points where the distance from the typhoon center point to the target location does not exceed the corresponding typhoon absolute scale, and their corresponding typhoon numbers and occurrence year numbers; the found path points, typhoon numbers, and occurrence year numbers are the objects to be examined next; 2) Count the typhoon numbers j (j = 1, 2,..., J) that affect the target location in the i-th year (i = 1, 2,..., I); 3) Obtain the typhoon path points k (k = 1, 2,..., K) of the j-th typhoon that affect the target location; 4) According to the typhoon state parameters of the path point k, use the typhoon wind field numerical simulation method to calculate the surface wind field of the target area, that is, the wind speed and wind direction, and record them; 5) If k < K, let k = k + 1 and return to step 4); if k = K, continue to the next step; 6) Count the wind field data caused by all path points k (k = 1, 2,..., K) of the j-th typhoon on the target area, find the maximum wind speed value and its corresponding wind direction, and record them as the extreme wind speed and the corresponding wind direction caused by the j-th typhoon; 7) If j < J, let j = j + 1 and return to step (3); if j = J, continue to the next step; 8) Count the extreme wind speed values and their wind directions caused by all typhoons j (j = 1, 2,..., J) in the i-th year, find the maximum extreme wind speed value and its corresponding wind direction, and record them as the annual extreme wind speed and the corresponding wind direction caused by the typhoons in the i-th year; 9) If i < I, let i = i + 1 and return to step (2); if i = I, continue to the next step; 10) Count the annual extreme wind speed values and their wind directions of all year numbers i (i = 1, 2,..., I) and record them, and then construct the joint probability model of typhoon wind speed and wind direction in the target area.

Images