A method and system for predicting reliability of a heavy haul railway overhead line system under thunderstorm weather

Abstract

The application discloses a heavy-load railway catenary reliability prediction method and system under thunder weather, and the method comprises the following steps: acquiring historical lightning strike data of the heavy-load railway catenary, and establishing a thunder generation model according to the historical lightning strike data of the heavy-load railway catenary; continuously dividing and sorting time periods of thunder weather at equal intervals, in each time period, using a density-based clustering algorithm to perform spatial clustering on lightning position information to obtain a thunder clustering cluster of each time period; combining thunder parameters to identify a region in which thunder occurs relatively densely in each clustering cluster, and then predicting a lightning range of a next time period through thunder clustering clusters of previous continuous time periods; combining the predicted lightning range and the thunder generation model to calculate a lightning strike failure probability of the heavy-load railway catenary; and calculating a reliability of the heavy-load railway catenary according to the lightning strike failure probability. The method has high usability and accuracy.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of traction power supply lightning protection and reliability analysis, in particular to a heavy haul railway catenary reliability prediction method and system under lightning weather. BACKGROUND

[0002] The catenary is the only power source of electric locomotives. If the catenary trips or breaks, it will cause traction power supply failure or even paralysis of the entire line. Lightning is a serious natural disaster. When lightning strikes, lightning overvoltage is generated on the contact line, which will cause the catenary to trip and even break in severe cases. According to statistics, the number of failures caused by lightning each year accounts for more than 50% of the total number of catenary failures. Therefore, lightning has become a major factor affecting the safety and reliability of catenary. Heavy haul railway equipment is mostly located in mountainous and hilly areas, and there are many iron ore mountains along the line, which further reduces the reliability of heavy haul railway catenary in rainy weather. Therefore, it is urgent to study the heavy haul railway catenary trip probability and reliability prediction technology under lightning weather, solve the problem of catenary reliability prediction under lightning conditions, and lay a theoretical foundation for heavy haul railway lightning weather early warning and maintenance strategy.

[0003] Existing overhead line lightning trip research can be divided into three categories: statistical, analytical and simulation. Statistical research: Doostan M., Alvehag K., Kankanala P. and other scholars have used binomial distribution model, time-varying failure rate model and exponential model to study the probability of failure of overhead transmission lines caused by lightning weather. Analytical research: The electrical geometry model (EGM) is a research and analysis method that is often widely used and is widely used to evaluate the lightning protection performance and reliability of transmission lines, distribution lines and catenary lines and other facilities. However, this model is mostly used for analysis of lightning trip probability in the case of annual average lightning, and lacks consideration of the randomness of the catenary lightning process, and cannot analyze the time-varying characteristics of the catenary lightning trip. Simulation research: Lucca G. used the Monte Carlo method to calculate the failure rate of overhead lines, Balijepalli N. used the Bootstrap method to model the lightning parameters, and then used the Monte Carlo method to evaluate the reliability indicators of the distribution network under lightning weather, and Chen J. also considered the randomness of the lightning process, combined with electromagnetic transient simulation and Monte Carlo method to calculate the lightning flashover probability of distribution lines. However, the simulation method has large amount of calculation and cannot achieve complete random sampling effect.

[0004] Therefore, the existing catenary lightning trip and reliability prediction method has the problems of low prediction accuracy and low feasibility. SUMMARY

[0005] The technical problem to be solved by the present application is that the existing overhead line system lightning fault tripping and reliability prediction method has the problems of low prediction accuracy and low feasibility. The present application aims to provide a heavy haul railway overhead line system reliability prediction method and system under lightning weather, based on modeling of lightning occurrence probability, clustering analysis of lightning position information and lightning parameters by DBSCAN algorithm, thereby predicting the lightning occurrence number and lightning range in the next time period by using two consecutive time lightning clusters; finally, combining the electrical geometric model, calculating the lightning tripping probability of a single group of overhead line system pillars, thereby obtaining the lightning tripping probability of the overhead line system powered by the heavy haul railway substation, and evaluating its reliability.

[0006] The present application is realized by the following technical solutions:

[0007] In the first aspect, the present application provides a heavy haul railway overhead line system reliability prediction method under lightning weather, which comprises:

[0008] Obtaining heavy haul railway overhead line system historical lightning data, and establishing a lightning occurrence model according to the heavy haul railway overhead line system historical lightning data;

[0009] Dividing and sorting the time periods of consecutive lightning weather at equal intervals, in each time period, using a density-based clustering algorithm to perform spatial clustering on lightning position information, to obtain lightning cluster of each time period;

[0010] Combining lightning parameters, identifying the areas with relatively dense lightning occurrence in each cluster, and then predicting the lightning range in the next time period by using the lightning clusters of the previous several consecutive time periods;

[0011] By establishing an electrical geometric model of the heavy haul railway overhead line system, combining the predicted lightning range and the lightning occurrence model, the lightning fault probability of the heavy haul railway overhead line system is calculated; and according to the lightning fault probability, the reliability of the heavy haul railway overhead line system is calculated.

[0012] Further, the lightning parameters include lightning occurrence time, lightning occurrence number, lightning longitude, lightning latitude, lightning current amplitude, ground flash density and lightning current steepness.

[0013] Further, the lightning occurrence model is established according to the heavy haul railway overhead line system historical lightning data, comprising:

[0014] Using a non-homogeneous Poisson process to replace the lightning occurrence events in the natural environment, the time interval between the lightning occurrence number and the lightning number in each time period is subject to a non-homogeneous Poisson distribution, and the lightning occurrence model is established.

[0015] Further, the expression of the lightning occurrence model is:

[0016]

[0017]

[0018] wherein P{N(t)=n} is the probability of n lightning strikes in the time interval (0,t]; N(t) is the number of lightning occurrences in the time interval (0,t]; V(t) is the cumulative lightning occurrence rate; n is the number of lightning strikes; and f(t) is the lightning occurrence probability at time t.

[0019] Further, the lightning cluster C tk of each time period is tk (X t ,Y t ,ε t ,N t ,P t ), wherein t is the time period to which the lightning cluster C tk belongs; X t and Y t are the longitude coordinate and the latitude coordinate of the center point of the lightning cluster C tk , respectively; ε t is the neighborhood radius of the lightning cluster; N t is the number of lightning strikes in the lightning cluster; and P t is the probability of lightning falling in the lightning cluster C tk in the time period t.

[0020] The longitude coordinate and the latitude coordinate of the center point of the lightning cluster C tk in the time period t are calculated according to the following formula:

[0021]

[0022] wherein x i and y i are the longitude coordinate and the latitude coordinate of the lightning strike point i in the lightning cluster, respectively.

[0023] The calculation formula of ε t is:

[0024] The calculation formula of P t is: wherein S represents the area of the lightning cluster C tk .

[0025] Further, the lightning strike range in the next time period is predicted by the lightning clusters in the previous two consecutive time periods, specifically as follows:

[0026] The lightning moving trend and speed are calculated according to the lightning cluster center point coordinates of the previous two time periods t-2 and t-1, and a first moving vector and a second moving speed vector are obtained;

[0027] A third moving speed vector of the lightning cluster center point from time period t to time period t+1 is calculated according to the first moving vector and the second moving speed vector;

[0028] A first change amount of the lightning stroke range from time period t-2 to time period t-1 is calculated, and a second change amount of the lightning stroke range from time period t-1 to time period t is calculated;

[0029] A third change amount of the lightning stroke range from time period t to time period t+1 is calculated according to the first change amount and the second change amount;

[0030] The lightning cluster neighborhood radius at time period t+1 is predicted according to the third change amount and the lightning cluster neighborhood radius at time period t;

[0031] The center coordinates of the lightning cluster at time period t+1 are predicted according to the third moving speed vector and the center coordinates of the lightning cluster at time period t, that is, the longitude and latitude of the lightning stroke range at time period t+1;

[0032] The number of lightning strokes of the lightning cluster at time period t+1 is predicted according to the lightning cumulative occurrence rate;

[0033] The lightning stroke density in the lightning cluster is calculated according to the lightning cluster neighborhood radius at time period t+1, the number of lightning strokes of the lightning cluster at time period t+1, and the ground flash density.

[0034] Further, the calculation formula of the lightning stroke density is:

[0035]

[0036] In the formula, P t+1 is the lightning stroke density in the lightning cluster at time period t+1; GFD is the ground flash density; GFD k is the lightning ground flash density in the kth lightning cluster, k=1, 2,..., m; m is the total number of lightning clusters at time period t+1; N t+1 represents the number of lightning strokes in the cluster predicted at time t+1.

[0037] Further, the calculation formula of the heavy haul railway overhead contact system lightning stroke failure probability is:

[0038]

[0039] Wherein, P is the contact network lightning trip-out probability in the m lightning clustering cluster in the time period t+1, as the heavy haul railway contact network lightning failure probability; m is the total number of lightning clustering clusters in the time period t+1; n k is the number of contact network support posts in the kth predicted lightning clustering cluster (i.e. lightning range); P l is the lightning occurrence probability calculated from the lightning cumulative occurrence rate V(t); P iP represents the lightning fall probability of the ith group of support posts, S′ i,k is the area of the predicted lightning range overlapping with the lightning attraction range of the ith group of support posts, S k is the area of the kth predicted lightning spatial cluster in the time period t+1; P i represents the lightning trip-out probability of the ith group of support posts; N k is the number of lightning falls in the kth lightning clustering cluster.

[0040] Further, according to the lightning failure probability and the railway reference reliability, the heavy haul railway contact network reliability is calculated, comprising:

[0041] According to the lightning failure probability, the heavy haul railway reference reliability is calculated;

[0042] According to the heavy haul railway reference reliability and the operation and maintenance contact network system reliability, the heavy haul railway contact network reliability is calculated.

[0043] In the second aspect, the present application further provides a heavy haul railway contact network reliability prediction system under lightning weather, which uses the heavy haul railway contact network reliability prediction method under lightning weather described above, and the system comprises:

[0044] An acquisition unit is configured to acquire heavy haul railway contact network historical lightning data;

[0045] A lightning occurrence model construction unit is configured to establish a lightning occurrence model according to the heavy haul railway contact network historical lightning data;

[0046] A lightning clustering unit is configured to divide and sort the time periods of continuous lightning weather at equal intervals, and in each time period, the spatial clustering of lightning fall location information is performed by using a density-based clustering algorithm to obtain the lightning clustering cluster of each time period;

[0047] A lightning fall range prediction unit is configured to identify the areas where lightning occurs more densely in each clustering cluster in combination with lightning parameters, and then predict the lightning fall range of the next time period through the lightning clustering clusters of the previous several continuous time periods;

[0048] The overhead contact line lightning stroke tripping probability calculation unit is used for calculating the lightning stroke fault probability of the heavy haul railway overhead contact line by establishing an electrical geometric model of the heavy haul railway overhead contact line and combining the predicted lightning range and the lightning occurrence model.

[0049] The overhead contact line reliability calculation unit is used for calculating the reliability of the heavy haul railway overhead contact line according to the lightning stroke fault probability.

[0050] Compared with the prior art, the present application has the following advantages and beneficial effects:

[0051] 1. The heavy haul railway overhead contact line reliability prediction method and system under lightning weather, first, the lightning occurrence probability model is established by combining the non-homogeneous Poisson process with the historical lightning data; then the basic time unit division is performed on the lightning data, and the clustering method is used to cluster the real-time lightning data of the adjacent three time periods; finally, the overhead contact line lightning stroke tripping probability and the overhead contact line reliability are calculated in combination with the electrical geometric model of the heavy haul railway overhead contact line. The method has high usability and accuracy, and the actual data verification analysis shows that the method can be used for actual overhead contact line lightning warning and improve the lightning fault defense level of heavy haul railway.

[0052] 2. The heavy haul railway overhead contact line reliability prediction method and system under lightning weather, a certain substation power supply arm along the heavy haul railway in China is selected for example research, and the real-time tripping probability and reliability of the overhead contact system under the influence of lightning weather are discussed and quantified. It is verified that the model applied to the actual overhead contact line has high feasibility in real-time calculation of lightning range and overhead contact line lightning stroke tripping probability. BRIEF DESCRIPTION OF DRAWINGS

[0053] The drawings described herein are used to provide further understanding of the embodiments of the present application, constitute a part of the present application, and do not constitute a limitation on the embodiments of the present application. In the drawings:

[0054] Figure 1 The flowchart of the heavy haul railway overhead contact line reliability prediction method under lightning weather is shown in the drawings;

[0055] Figure 2 The central point and density display schematic diagram is shown in the drawings;

[0056] Figure 3 The specific steps of the noise application space clustering algorithm based on density are shown in the drawings;

[0057] Figure 4 The lightning occurrence position statistical result graph of a certain power supply arm near the railway in 2021 is shown in the drawings;

[0058] Figure 5 The lightning occurrence probability density function graph based on the non-homogeneous Poisson process is shown in the drawings;

[0059] Figure 6 The application is a heavy load railway catenary reliability prediction system structure diagram under thunder weather. DETAILED DESCRIPTION

[0060] In order to make the purpose, technical scheme and advantages of the application clearer, the application is further described in detail below in combination with examples and drawings, and the schematic embodiment of the application and the description thereof are only used to explain the application, and do not limit the application.

[0061] Based on the existing catenary lightning fault tripping and reliability prediction method, there are problems of low prediction accuracy and low feasibility. The application designs a heavy load railway catenary reliability prediction method and system under thunder weather,

[0062] On the basis of modeling the lightning occurrence probability, the lightning position information and lightning parameters are analyzed by the DBSCAN algorithm, so as to predict the lightning occurrence times and lightning range in the next time period by using a plurality of (such as two) continuous time lightning clusters; finally, the lightning tripping probability of a single group of catenary pillars is calculated by combining the electrical geometric model of the catenary, so as to obtain the lightning tripping probability of the catenary powered by the heavy load railway substation, and evaluate the reliability. The method has high availability and accuracy, and the actual data verification analysis shows that the method can be used for actual catenary lightning early warning, and improve the lightning fault defense level of heavy load railway.

[0063] The application mainly includes the following parts:

[0064] First, lightning occurrence model

[0065] Considering that the existing lightning tripping calculation model only uses historical thunderstorm data to calculate the annual lightning tripping rate in combination with the electrical geometric model, however, the influence of real-time changing lightning data on the lightning tripping probability and reliability of the heavy load railway catenary is ignored. The application firstly fits the distribution function of lightning parameters (ground flash density, lightning current and lightning current steepness, etc.) by using historical lightning data, and establishes a lightning occurrence model by using a non-homogeneous Poisson process, so as to facilitate subsequent analysis of the lightning tripping probability and reliability of the catenary system.

[0066] Second, lightning range prediction based on clustering algorithm

[0067] Since lightning weather has seasonality, basically lightning weather occurs more frequently in summer every year, therefore the present application selects time period in which lightning weather occurs frequently to research the catenary. When lightning weather occurs continuously, the time period is divided and sorted in equal interval, in each time interval, the longitude and latitude information of lightning can be clustered in space by using density-based clustering algorithm, in combination with lightning parameters such as lightning occurrence time, occurrence frequency and lightning current amplitude, the area in which lightning occurs more densely is identified, and then the lightning range is predicted.

[0068] Third, heavy haul railway catenary lightning fault probability calculation and reliability prediction

[0069] The electrical geometric model is a kind of electrical geometric analysis model which combines lightning discharge characteristics and other lightning index parameters with the structure and size of the actual researched railway catenary. Therefore, the lightning fault probability of heavy haul railway catenary can be calculated by establishing the electrical geometric model of heavy haul railway catenary in combination with the predicted lightning range, so as to calculate the reliability of catenary and realize the reliability prediction of heavy haul railway catenary under lightning weather.

[0070] Embodiment 1

[0071] As shown in Figure 1 , the present application is a kind of reliability prediction method of heavy haul railway catenary under lightning weather, which comprises:

[0072] Step 1, obtaining the historical lightning data of heavy haul railway catenary, including lightning weather occurrence data and catenary trip-out fault history record data under the influence of lightning weather.

[0073] Step 2, establishing a lightning occurrence model according to the historical lightning data of heavy haul railway catenary, including:

[0074] The present application uses a non-homogeneous Poisson process to replace the lightning occurrence event in natural environment, so that the time interval between the number of lightning in each time and the number of lightning obeys non-homogeneous Poisson distribution, and establishes a lightning occurrence model.

[0075] Specifically, the expression of the lightning occurrence model is:

[0076]

[0077]

[0078] In the formula, P{N(t)=n} is the probability of n lightning in time (0,t]; N(t) is the number of lightning in time interval (0,t]; V(t) is the cumulative lightning rate; n is the number of lightning; f(t) is the lightning occurrence probability at time t.

[0079] Therefore, the expectation of the number of lightning occurrences in the time interval from t to t+Δt is:

[0080] E[N(t+Δt)-N(t)]=V(t+Δt)-V(t) (3)

[0081] Step 3, divide the continuous lightning weather period into equal intervals and sort them, in each time interval, use the density-based clustering algorithm to cluster the lightning location information in space, and get the lightning clustering cluster of each time interval;

[0082] (1) DBSCAN algorithm

[0083] The DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm is considered as a high-density object space obtained by dividing a low-density space, a set of similar points is assigned to a cluster, and data points in a high-density area are grouped and called clusters, and data points not belonging to any cluster are designated as noise.

[0084] The basic principle of the DBSCAN algorithm is as follows: as shown in Figure 2 , p and q represent spatial entities in the object set Γ, ε represents the maximum Euclidean distance for two points to be in the same neighborhood, and N ε (p) represents the set of spatial entities within a distance of ε from entity p, and Entity min (p) represents the minimum number of spatial entities in the neighborhood circle centered on entity p; |N ε (p) | is the number of spatial entities in the neighborhood of p, if |N ε (p) |≥Entity min (p), then the spatial entity p is a center point. If q∈N ε (p) and spatial entity p is a center point, then p is directly density reachable for q, and this property is transitive.

[0085] Based on the above DBSCAN algorithm principle, the specific steps of the density-based noise application spatial clustering algorithm can be obtained as shown in the flowchart Figure 3 .

[0086] (2) Lightning clustering based on DBSCAN algorithm

[0087] The thunder weather often lasts for half an hour to several hours, and the thunder lasts for about 5 minutes, so the present application takes the lightning in 5 minutes as the thunder in the same time period. After dividing the thunder occurrence data according to the time period, the thunder monitoring information in the same time period is selected for cluster analysis. According to the above principle, for any lightning position p, the following formula is known:

[0088] N ε ={dist(q,p)≤ε} (4)

[0089] Wherein, epsilon represents the neighborhood radius of the thunder cluster C, which and Entity min (p) are set according to actual conditions in application, and dist(q,p) represents the distance between the lightning q and the point p.

[0090] The continuous thunder weather period is divided and sorted at equal intervals, in each time period, the lightning position information is spatially clustered by using the density-based clustering algorithm, and the thunder cluster C tk of each time period is obtained. tk (X t ,Y t ,epsilon t ,N t ,P t ), wherein t is the time period to which the thunder cluster C tk belongs; X t and Y t are the longitude coordinate and latitude coordinate of the center point of the thunder cluster C tk , respectively; epsilon t is the neighborhood radius of the thunder cluster; N t is the number of lightning in the thunder cluster; and P t is the probability of lightning falling in the thunder cluster C tk in the time period t.

[0091] The longitude coordinate and latitude coordinate of the center point of the thunder cluster C tk in the time period t are calculated by the following formula:

[0092]

[0093] In the formula, x i and y i are the longitude coordinate and latitude coordinate of the lightning point i in the thunder cluster, respectively.

[0094] The neighborhood radius epsilon t of the thunder cluster is calculated by the following formula:

[0095]

[0096] The density P of lightning falling in the cluster cluster in the time period t t The calculation formula is:

[0097]

[0098] Wherein, S represents the area of lightning cluster C tk .

[0099] Step 4, combined with lightning parameters, the area where lightning occurs more densely in each cluster cluster is identified, and then the lightning cluster in the previous several continuous time periods is used to predict the lightning range in the next time period;

[0100] On the basis of the lightning cluster obtained in step 3, the lightning cluster in several continuous time periods is used to predict the lightning range in the next time period.

[0101] Specifically, the method calculates the moving trend and speed of the center point coordinates of the lightning cluster in the previous two continuous time periods t-2 and t-1, obtains the corresponding first moving vector V Ct-2 , second moving speed vector V Ct-1 ; here, the vector result is represented as:

[0102]

[0103]

[0104] In the formula, (X (t-2) , Y (t-2) ), (X (t-1) , Y (t-1) ) and (X t , Y t ) are the center point coordinates of the lightning cluster in time period t-2, time period t-1 and time period t respectively.

[0105] Calculate the first change amount λ t-2 of lightning stroke range from time period t-2 to time period t-1, and calculate the second change amount λ t-1 of lightning stroke range from time period t-1 to time period t;

[0106]

[0107] According to the first moving vector V Ct-2 and the second moving speed vector V Ct-1 , the third moving speed vector V Ct of the center point of the lightning cluster from time period t to time period t+1 is calculated;

[0108]

[0109] According to the first change amount λ t-2 and the second change amount λ t-1 , a third change amount λ t of the lightning strike range from the time period t to the time period t+1 is calculated;

[0110]

[0111] According to the third moving speed vector V Ct and the center coordinates (X t , Y t ) of the lightning cluster at the time period t, the center coordinates (X t+1 , Y t+1 ) of the lightning cluster at the time period t+1 are predicted, i.e. the longitude and latitude of the lightning strike range at the time period t+1;

[0112] (X t+1 , Y t+1 ) = (X t , Y t ) + V Ct · Δt (12)

[0113] wherein Δt is a time interval divided according to the lightning duration.

[0114] According to the third change amount λ t and the neighborhood radius ε t of the lightning cluster at the time period t, the neighborhood radius ε t+1 of the lightning cluster at the time period t+1 is predicted;

[0115] ε t+1 = ε t (1 + λ t ) (13)

[0116] In combination with formula (2), the lightning strike number N t+1 of the lightning cluster at the time period t+1 is predicted according to the lightning cumulative occurrence rate;

[0117]

[0118] According to the neighborhood radius ε t+1 of the lightning cluster at the time period t+1, the lightning strike number N t+1 of the lightning cluster at the time period t+1 and the ground flash density GFD, the lightning strike density P t+1 in the lightning cluster is calculated, and the calculation formula is:

[0119]

[0120] In the formula, P t+1is the lightning ground flash density in the kth lightning cluster, k = 1, 2,..., m; m is the total number of lightning clusters at time period t + 1; N k is the lightning ground flash density in the kth lightning cluster, k = 1, 2,..., m; m is the total number of lightning clusters at time period t + 1; N t+1 represents the number of lightning in the cluster at time t + 1.

[0121] Formula (15) calculates the lightning weight of each lightning cluster by calculating the hourly ground flash density data obtained from the China Meteorological Administration, so as to obtain the lightning density in a cluster.

[0122] The lightning current size of each lightning in the lightning cluster is determined by the lightning current probability distribution f I (I) is sampled.

[0123] Step 5, by establishing an electrical geometric model of the heavy haul railway overhead contact system, combining the predicted lightning range and the lightning occurrence model, the lightning fault probability of the heavy haul railway overhead contact system is calculated; according to the lightning fault probability, the reliability of the heavy haul railway overhead contact system is calculated.

[0124] (1) Contact net lightning trip-out probability

[0125] The contact net lightning trip-out often needs to consider two cases: the support, feeder, and return line are hit or the dropper, messenger wire, and contact line are hit. If the contact net is hit by lightning, whether the contact net trips depends on whether the lightning current exceeds the lightning withstand level of the contact net.

[0126] When the support, feeder, and return line are hit, the lightning withstand level is:

[0127]

[0128]

[0129] Where, R is the grounding resistance of the support, L is the equivalent inductance of the support, U F50％ is the flashover voltage of the feeder, h T is the average height of the messenger wire, h F is the average height of the feeder, τ is the time before the lightning wave, generally, the value is 2.65us. k0 is the geometric coupling coefficient between the feeder and the messenger wire, k1 is the corona correction coefficient, if it is single lightning conductor, it is 1.1-1.2, if it is double lightning conductor, it is 1.15-1.25, L g is the parallel inductance of the protection ground wire on both sides of the support.

[0130] In this case, the trip-out probability is:

[0131]

[0132] In the formula, fI (I) is a lightning current probability density distribution function.

[0133] When the dropper, contact wire, and messenger wire are hit, the lightning withstand level is as follows:

[0134]

[0135] wherein U 50％ is the contact wire flashover voltage, k0 is the geometric coupling coefficient between the feeder and the messenger wire, and k1 is the corona correction coefficient, which is taken as 1.1-1.2 if it is a single lightning conductor and 1.15-1.25 if it is a double lightning conductor.

[0136] In this case, the trip-out probability is:

[0137]

[0138] As described above, for the ith group of pillars (i = 1, 2,..., n k , n k is the number of contact net pillars in the kth predicted lightning cluster (i.e., lightning range), if the predicted lightning range and the lightning range of the ith group of pillars overlap, the probability of contact net lightning trip-out is as follows:

[0139] P i = ξ(P a P1+P b P2) (21)

[0140] wherein ξ represents the arc establishment rate, P a represents the probability of lightning strike on the pillars, feeder, and return wire, and P b represents the probability of lightning strike on the dropper, contact wire, and messenger wire.

[0141] Therefore, the contact net lightning trip-out probability in the m lightning clusters at time period t+1, i.e., the calculation formula of the heavy-haul railway contact net lightning failure probability, can be obtained as follows:

[0142]

[0143] wherein P is the contact net lightning trip-out probability in the m lightning clusters at time period t+1, which is the heavy-haul railway contact net lightning failure probability; m is the total number of lightning clusters at time period t+1; n k is the number of contact net pillars in the kth predicted lightning cluster (i.e., lightning range); P l is the lightning occurrence probability obtained from formula (2); P iP represents the lightning probability of the ith group of pillars, S′ i,k is the area of the overlapping part of the predicted lightning range and the lightning range of the ith group of pillars, and S kP is the area of the kth lightning spatial cluster in the predicted time period t+1; P i N represents the lightning trip-out probability of the ith group of pillars; N k Nk represents the number of lightning strikes in the kth lightning cluster.

[0144] Formula (22) is the lightning trip-out probability prediction model of the present application.

[0145] (2) Reliability prediction of heavy haul railway under lightning weather

[0146] If the reliability of the heavy haul railway is to be calculated, first, the reference reliability of the heavy haul railway is calculated according to the lightning failure probability; the reference reliability function is as follows:

[0147]

[0148] Secondly, the reliability of the overhead contact system of the heavy haul railway is calculated according to the reference reliability of the heavy haul railway and the operation and maintenance reliability of the overhead contact system. The present application analyzes the reliability of the overhead contact system of the railway under lightning weather, and the railway reliability function can be obtained as follows:

[0149] R(t) = R m (t)R base (t) (24)

[0150] Wherein, R m (t) represents the reliability of the heavy haul railway overhead contact system under the condition of routine maintenance and replacement of parts.

[0151] The present application analyzes the occurrence data of lightning weather and the historical record data of the trip-out failure of the overhead contact system under the influence of lightning weather, and innovatively proposes the lightning trip-out probability prediction and reliability model of the overhead contact system. The present application selects a power supply arm of a certain substation along the heavy haul railway in China for example research, discusses and quantifies the real-time trip-out probability and reliability of the overhead contact system under the influence of lightning weather. It is verified that the model has high feasibility in the actual overhead contact system, whether in the real-time calculation of the lightning range or the lightning trip-out probability of the overhead contact system.

[0152] In specific implementation, the overhead contact system along the Shuozhou-Huanghuagang heavy haul railway is implemented, as follows:

[0153] First, the lightning data along the heavy haul railway is collected. The lightning occurrence positions around the power supply arm of a certain substation along the railway are counted, and the results are shown in Table 1. Figure 4

[0154] ​According to the statistical analysis of the obtained historical lightning data, the lightning activity is fitted using the non-homogeneous Poisson model proposed in step 2, and a lightning occurrence model is established for each hour of the year. It can be seen that the lightning occurrence density function subject to the non-homogeneous Poisson distribution is as shown in formula (21). Figure 5

[0155] Secondly, according to the lightning trip-out prediction model proposed above, based on the lightning occurrence in the last three time periods, the prediction of the lightning range in the next five minutes can be obtained, and the lightning trip-out probability and reliability of the heavy-load railway overhead contact system can be obtained by combining the lightning trip-out probability of the overhead contact system given by formula (22). Next, the present application is demonstrated from the micro and macro aspects:

[0156] Example 1: The lightning parameters in a certain area of the railway from 9:00 to 10:00 on July 14, 2022 are shown in Table 1. Table 2 lists the prediction results of the lightning trip-out probability of the overhead contact system in the time period with high trip-out probability.

[0157] Table 1 Lightning data in the area near a certain substation of the railway from 9:00 to 10:00 on July 14, 2022 (part)

[0158]

[0159]

[0160] Table 2 Prediction results of high risk of lightning trip-out of overhead contact system from 9:00 to 10:00 on July 14, 2022

[0161] Forecast time Site Lightning trip rate Number of tower groups within the forecast lightning range 2022-07-14 09:10 Site 1 - Site 2 7.34％ 1044 2022-07-14 09:32 Site 1 - Site 2 3.46％ 503 2022-07-14 09:40 Site 1 - Site 2 6.39％ 779

[0162] According to the historical trip-out report records, the lightning trip-out accident occurred at 9:11 on July 14, 2022, which is consistent with the prediction results.

[0163] Example 2: For the lightning trip-out faults of the heavy-load railway overhead contact system from 2020 to 2022, a total of 15 lightning trip-out times of the overhead contact system occurred. The lightning trip-out probability prediction model (formula (22)) proposed in the present application is applied to analyze the results, and Table 3 is the prediction calculation results of the lightning trip-out probability obtained by using the model.

[0164] Table 3 Prediction of lightning trip-out results of overhead contact system of heavy-load railway line from 2020 to 2022

[0165]

[0166]

[0167] ​Finally, the reliability prediction analysis of the catenary of the heavy haul railway under the influence of thunder and lightning weather on July 14, 2022 from 9:00 to 10:00 can be obtained according to the catenary reliability model proposed in the foregoing, as shown in Table 4.

[0168] Table 4 Reliability prediction of catenary of heavy haul railway line from 9:00 to 10:00 on July 14, 2022

[0169]

[0170] As can be seen from Table 4, the catenary reliability is the lowest between 9:10 and 9:15 in this time period, and the trip-out accident caused by lightning strike occurred at 9:11 on July 14, 2022 according to the fault history record data, which is consistent with the analysis result, so it can be obtained that the catenary reliability model of the heavy haul railway under the influence of thunder and lightning weather proposed in this paper has high availability and accuracy.

[0171] Embodiment 2

[0172] As Figure 6 shown, the difference between this embodiment and Embodiment 1 is that this embodiment provides a system for predicting the reliability of the catenary of the heavy haul railway under the influence of thunder and lightning weather, which uses the method for predicting the reliability of the catenary of the heavy haul railway under the influence of thunder and lightning weather in Embodiment 1, and the system corresponds to the method for predicting the reliability of the catenary of the heavy haul railway under the influence of thunder and lightning weather in Embodiment 1 one by one. The system comprises:

[0173] An acquisition unit is configured to acquire historical lightning strike data of the catenary of the heavy haul railway.

[0174] A lightning occurrence model construction unit is configured to establish a lightning occurrence model according to the historical lightning strike data of the catenary of the heavy haul railway in combination with a non-homogeneous Poisson process.

[0175] A lightning clustering unit is configured to divide and sort the time periods of continuous lightning weather at equal intervals, and in each time period, a spatial clustering algorithm based on density is used to cluster the lightning location information to obtain lightning clustering clusters in each time period.

[0176] A lightning range prediction unit is configured to identify the areas where lightning occurs more densely in each clustering cluster in combination with lightning parameters, and then predict the lightning range in the next time period through the lightning clustering clusters in the previous continuous time periods.

[0177] A catenary lightning trip-out probability calculation unit is configured to calculate the lightning fault probability of the catenary of the heavy haul railway in combination with the predicted lightning range and the lightning occurrence model.

[0178] A catenary reliability calculation unit is configured to calculate the reliability of the catenary of the heavy haul railway according to the lightning fault probability.

[0179] The execution process of each unit is performed according to the process steps of the heavy-load railway overhead line system reliability prediction method in thunder weather described in Embodiment 1, and the embodiment will not be described again.

[0180] The system firstly establishes a thunder generation probability model by combining historical thunder data with a non-homogeneous Poisson process; then, the thunder data is divided into basic time units, and clustering is performed on real-time thunder data in adjacent three time periods; finally, the overhead line system thunder trip probability and the overhead line system reliability are calculated by combining an electrical geometric model of the heavy-load railway overhead line system. Actual data verification and analysis show that the method can be used for actual overhead line system thunder early warning and improve the heavy-load railway thunder fault defense level.

[0181] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage, etc.) containing computer-usable program code.

[0182] The present application is described with reference to flowcharts and / or block diagrams according to the methods, devices (systems), and computer program products of the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to produce a machine, so that the instructions executed by the computer or other programmable data processing devices produce a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks. Figure 1 The devices that implement the functions specified in one or more flows and / or blocks.

[0183] These computer program instructions can also be stored in a computer-readable memory that can guide the computer or other programmable data processing devices to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including instruction devices that implement the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks. Figure 1 The devices that implement the functions specified in one or more flows and / or blocks.

[0184] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0185] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for predicting the reliability of a heavy haul railway overhead contact system in thunderstorm weather, characterized in that, The method comprises: Obtaining historical lightning strike data of the heavy haul railway overhead contact system, and establishing a lightning occurrence model according to the historical lightning strike data of the heavy haul railway overhead contact system; Dividing and sorting time periods of continuous lightning weather into equal intervals, and in each time period, using a density-based clustering algorithm to perform spatial clustering on lightning location information to obtain lightning clustering clusters in each time period; Combining lightning parameters, identifying regions with dense lightning occurrence in each clustering cluster, and predicting a lightning range in a next time period through lightning clustering clusters in previous continuous time periods; Combining the predicted lightning range and the lightning occurrence model, calculating a lightning strike failure probability of the heavy haul railway overhead contact system, and calculating a reliability of the heavy haul railway overhead contact system according to the lightning strike failure probability; Lightning clusters C for each time period tk Represented as C tk (X t ,Y t ,ε t N t ,P t ), where t is C tk Time period; X t Y t They are lightning clusters C tk The longitude and latitude coordinates of the center point; ε t N represents the neighborhood radius of the lightning cluster; t P represents the number of lightning strikes in the lightning cluster; t For time period t, lightning falls on lightning cluster C. tk The probability of it; Lightning cluster C tk The longitude coordinate and the latitude coordinate of the center point at the time period t are calculated by the following formulas: where x i ,y i are the longitude and latitude coordinates of the lightning point i in the lightning cluster, respectively. ε t The calculation formula is: P t The calculation formula is: Wherein, S represents the area of the lightning cluster C tk ​ The lightning range in the next time period is predicted through lightning clustering clusters in the previous two continuous time periods, and specifically: Calculating a lightning movement trend and a speed through center point coordinates of lightning clustering clusters in the previous two time periods t-2 and t-1 to obtain a corresponding first movement vector and a second movement speed vector; According to the first movement vector and the second movement speed vector, a third movement speed vector of the center point of the lightning clustering cluster from the time period t to the time period t+1 is calculated; A first change amount of the lightning strike range from the time period t-2 to the time period t-1 is calculated, and a second change amount of the lightning strike range from the time period t-1 to the time period t is calculated; According to the first change amount and the second change amount, a third change amount of the lightning strike range from the time period t to the time period t+1 is calculated; According to the third change amount and a neighborhood radius of the lightning clustering cluster on the time period t, the neighborhood radius of the lightning clustering cluster on the time period t+1 is predicted; According to the third movement speed vector and the center coordinates of the lightning clustering cluster on the time period t, the center coordinates of the lightning clustering cluster on the time period t+1 are predicted, that is, the longitude and latitude of the lightning range on the time period t+1; According to a lightning cumulative occurrence rate, the number of lightning strikes of the lightning clustering cluster on the time period t+1 is predicted; According to the neighborhood radius of the lightning clustering cluster on the time period t+1, the number of lightning strikes of the lightning clustering cluster on the time period t+1, and the ground flash density, a lightning density in the lightning clustering cluster is calculated; Combining the predicted lightning range and the lightning occurrence model, a lightning trip-out probability prediction model is used to calculate the lightning strike failure probability of the heavy haul railway overhead contact system. The calculation formula of the lightning trip-out probability prediction model is: Wherein, P is the contact net lightning trip-out probability in the m lightning clustering cluster in the time period t+1, as the heavy load railway contact net lightning failure probability; m is the total number of lightning clustering cluster in the time period t+1; n k is the number of contact net support in the kth lightning clustering cluster in the prediction; P l is the lightning occurrence probability; P iP represents the lightning fall probability of the ith group of support, S′ i,k is the area of the predicted lightning strike range and the lightning attracting range of the ith group of support, S k is the area of the kth lightning spatial cluster in the time period t+1 in the prediction; P i represents the lightning trip-out probability of the ith group of support; N k is the number of lightning fall in the kth lightning clustering cluster.

2. The method according to claim 1, wherein, The lightning parameters include lightning occurrence time, lightning occurrence number, lightning longitude, lightning latitude, lightning current amplitude, ground flash density, and lightning current steepness.

3. The method according to claim 1, wherein, According to the historical lightning strike data of the heavy haul railway overhead contact system, a lightning occurrence model is established, comprising: Using a non-homogeneous Poisson process to replace lightning occurrence events in a natural environment, and making the time interval between the number of lightning strikes and the number of lightning in each time period subject to a non-homogeneous Poisson distribution to establish the lightning occurrence model.

4. The method according to claim 3, wherein, Wherein, The expression of the lightning occurrence model is: In the formula, P{N(t)=n} is the probability of n lightning strikes in time (0,t]; N(t) is the number of lightning occurrences in the time interval (0,t]; V(t) is the lightning cumulative occurrence rate; n is the number of lightning strikes; and f(t) is the lightning occurrence probability at time t.

5. The method according to claim 1, wherein, The formula for calculating the lightning density is: where P t+1 is the lightning density in the lightning cluster at time period t+1; GFD is the ground flash density; GFD k is the lightning ground flash density in the kth lightning cluster, k = 1, 2, …, m; m is the total number of lightning clusters at time period t+1; N t+1 represents the predicted number of lightning times in the cluster at time t+1.

6. The method according to claim 1, wherein, According to the lightning failure probability and the railway benchmark reliability, the reliability of the heavy-haul railway overhead contact system is calculated, including: According to the lightning failure probability, the benchmark reliability of the heavy-haul railway is calculated; According to the benchmark reliability of the heavy-haul railway and the operation and maintenance reliability of the overhead contact system, the reliability of the heavy-haul railway overhead contact system is calculated.

7. A lightning weather under heavy load railway catenary reliability prediction system, characterized by, The system uses a heavy-haul railway overhead contact system reliability prediction method under lightning weather as claimed in any one of claims 1 to 6; the system includes: An acquisition unit for acquiring historical lightning data of the heavy-haul railway overhead contact system; A lightning occurrence model construction unit for establishing a lightning occurrence model according to the historical lightning data of the heavy-haul railway overhead contact system; A lightning clustering unit for equally dividing and sorting the time periods when lightning weather occurs continuously, and for using a density-based clustering algorithm to cluster the lightning location information in space in each time period to obtain lightning clustering clusters in each time period; A lightning range prediction unit for identifying the areas where lightning occurs densely in each clustering cluster in combination with lightning parameters, and for predicting the lightning range in the next time period through the lightning clustering clusters in the previous several continuous time periods; An overhead contact system lightning trip-out probability calculation unit for calculating the lightning failure probability of the heavy-haul railway overhead contact system in combination with the predicted lightning range and the lightning occurrence model; An overhead contact system reliability calculation unit for calculating the reliability of the heavy-haul railway overhead contact system according to the lightning failure probability.

Images