CN109948204B - Bayesian network-based dynamic risk analysis method for high-speed rail overhead line system - Google Patents
Bayesian network-based dynamic risk analysis method for high-speed rail overhead line system Download PDFInfo
- Publication number
- CN109948204B CN109948204B CN201910165009.0A CN201910165009A CN109948204B CN 109948204 B CN109948204 B CN 109948204B CN 201910165009 A CN201910165009 A CN 201910165009A CN 109948204 B CN109948204 B CN 109948204B
- Authority
- CN
- China
- Prior art keywords
- risk
- dynamic
- speed rail
- passenger flow
- station
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Abstract
The invention provides a Bayesian network-based dynamic risk analysis method for a high-speed rail contact network, which comprises the steps of establishing a risk propagation chain model of insulator flashover of the high-speed rail contact network, carrying out dynamic mathematical modeling on the risk chain model by using characteristic quantity-dynamic probability, and calculating the dynamic mathematical modeling in the Bayesian network to obtain risk occurrence probability; determining the dynamic severity of risks in two dimensions of time and space by using the passenger flow intensity index, calculating the severity of risk consequences caused by train delay at different times by using basic time granularity and stations as basic units for obtaining real-time monitoring data and using a risk propagation path and a propagation process as the principle of calculating the passenger flow intensity index; and combining the dynamic probability with the severity, and obtaining the dynamic risk level of the risk chain according to the risk assessment matrix.
Description
Technical Field
The invention belongs to the technical field of risk analysis of a high-speed rail power supply system, and particularly relates to a dynamic risk analysis method of a high-speed rail contact network based on a Bayesian network.
Background
In recent years, with the development and popularization of high-speed railways in China, people pay attention to the safety problems of the high-speed railways. According to the investigation and statistics results of the railway department on the electrified railway operation accidents of 2 kilometers all over the country, the accidents of train operation interruption caused by the power supply system faults in China account for more than half of all the accidents by 2016 years. In a high-speed rail power supply system, a contact network is a special power transmission line erected overhead along a high-speed electrified railway and is responsible for supplying power to an electric locomotive. The contact net is erected in the open along a railway line, and the working environment is severe; meanwhile, the contact net is used as a special power supply device, the contact net is connected by sliding the pantograph and the contact line in a contact manner, and under the condition of high-speed operation, the service life of elements of the contact net is easily reduced due to mutual abrasion of the pantograph and the contact line; in addition, as the contact net is not standby, once the contact net is damaged, the high-speed rail can not normally operate and even safety accidents happen. Therefore, the method for analyzing and evaluating the risks of the contact network is deeply researched and reasonably formulated, so that the accident rate of the contact network can be reduced, and the loss caused by the faults of the contact network can be reduced.
The research process for safety evaluation of high-speed rail goes through two main stages of probability evaluation and risk evaluation. In probability evaluation, a reliability analysis method based on probability is mostly used as a basis of a security evaluation method. For example, the reliability of the high-speed rail power supply system is analyzed by using a fault tree method. Reliability evaluation is carried out on the contact network in Bayesian network-based contact network operation reliability evaluation by Wangbei, a Bayesian network model of the contact network system is established, and the weakest link influencing the reliable operation of the contact network is obtained. The risk assessment is based on probabilistic assessment, and meanwhile, the consequences caused by the occurrence of the event are considered.
The high-speed railway contact network system is the main framework of railway electrification engineering and takes charge of the important task of supplying electric power to electric locomotives. As shown in fig. 1, the catenary mainly includes parts such as a contact line, a catenary, an insulator, a dropper, and a cantilever, and is erected on a support by a support device to transmit electric energy obtained from traction power transformation to an electric locomotive. In a contact network system, an insulator is used as one of the most used devices in the contact network, and plays a role in connecting a charged part of the contact network, a supporting device and other devices connected with the ground so as to ensure the electrical insulation between a charged body and the ground. The contact network system has numerous devices which are arranged in an up-down mode and are usually erected outdoors, the change of the external environment causes great interference to the normal operation of the contact network system, and one of the common and serious accidents is a pollution flashover accident of a contact network insulator under the operating voltage.
The method is characterized in that the mechanism of insulator pollution discharge is deeply researched in the establishment of a contact network insulator pollution degree prediction model by Huai Mengqi and the like, and the relation between the leakage current of the insulator and the environmental temperature and humidity is obtained through an artificial pollution test; the scenic soldiers provide an evaluation model of insulator pollution conditions through experiments in 'contact network insulator leakage current characteristics and state detection research' and provide reference basis for contact network insulator pollution removal cleaning and state evaluation in actual operation. The change of the operation environment can cause the flashover probability in the operation of a high-speed rail to change correspondingly, and although physical quantities such as pollution degree, leakage current and the like are extracted as characteristic quantities in the existing research aiming at the flashover of the high-speed rail contact network insulator, the attention to the dynamic characteristic quantity-dynamic fault probability is less. In the evaluation of the operation reliability of the power system based on the real-time operation state, the Sun-Yuan Chapter and the like, a dynamic outage probability change trend is assumed on the basis of a traditional outage model of a power system circuit according to the real-time operation condition of the circuit, and the model is verified through data; zhoutonhua et al, in "real-time reliability assessment and prediction technology for engineering systems" indicate that in actual engineering, real-time reliability indexes can be calculated by monitoring real-time changes of environmental variables and state variables.
From the above, although many scholars have studied on the operation safety of the contact network system, the research contents are less concerned about the real-time influence of the real-time change of the external environment on the operation of the system and the risk chain propagation process based on the failure mechanism. In addition, after a high-speed rail contact network fails, passenger flow congestion of stations and lines can be caused, and great potential safety hazards are brought, but current research only considers the possible consequences in a power supply system after the failure, and related evaluation is also lacked for the passenger flow safety risk condition brought by chain transmission of flashover risks.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides a Bayesian network-based dynamic risk analysis method for a high-speed rail contact network. Therefore, the set relevant threshold value can be compared with the change of the characteristic quantity in actual operation, corresponding measures are taken to block a risk propagation chain in a proper propagation link, and the risk of serious flashover accidents caused by the fact that the contact network is influenced by weather is reduced.
The specific technical scheme of the invention is as follows:
the invention provides a Bayesian network-based dynamic risk analysis method for a high-speed rail catenary, which comprises the following steps of:
s1: establishing a risk propagation chain model of the high-speed rail contact network insulator flashover, solving dynamic probabilities corresponding to all characteristic quantities in the risk propagation chain model, and substituting all the dynamic probabilities into the Bayesian network to obtain risk occurrence probabilities;
s2: calculating passenger flow intensity indexes according to risk propagation paths and propagation processes in the risk propagation chain model, and calculating the severity of risk consequences caused by train delay at different times;
s3: the Risk assessment matrix Risk shown in the formula (10) is given based on the Risk occurrence probability obtained in step S1 and the severity obtained in step S2,
Risk＝f(p，c) (10)
p is the risk occurrence probability and c is the severity;
s4: and evaluating the dynamic risk of the high-speed rail overhead line system according to the risk evaluation matrix.
In a further improvement, each characteristic quantity in the risk propagation chain model comprises pollution degree, leakage current and short-circuit current.
In a further refinement, the severity Yzcd is characterized by formula (6):
Yzcd＝K×t (6)
k represents the line passenger flow intensity index, and t is train delay time.
In a further refinement, the risk assessment matrix of step S3 is as follows:
in a further improvement, the dynamic probability includes a disconnection probability P (E | C) calculated by equation (5):
i is the effective value of the short-circuit current passing through the contact line; i is0Fusing current for the wire; t is tscThe duration of the short-circuit current is S is the sectional area of the contact line; a. thewFor parameters relating to the working temperature and material of the contact line, AhIs the parameter related to the highest temperature of the contact line and the material of the contact line in short circuit.
In a further improvement, the passenger flow intensity index is derived from the congestion degree and the congestion range of the line.
Further improvement, congestion level at basic time granularity
Calculated according to equation (7):
in the formula, alpha11、α12…α1iThe weights of the 1 st station, the 2 nd station and the i th station are respectively; mu.s1、μ2To muiPassenger flow density indexes of 1 st station, 2 th station and i th station respectively;
the degree of crowding of the station for 1min under the condition of normal running of the train is shown.
Further improvement, the crowding range theta at the basic time granularityCrowding for 1minCalculated from equation (11):
in the formula phiareaThe passenger flow density value threshold value of each basic observation unit is given for each station; phijThe passenger flow density value of the jth basic observation unit of the station; mjThe number of the basic observation units exceeding the passenger flow density value threshold value is N, and the N is the total number of the basic observation units.
Wherein, basic observation unit includes platform, access & exit and passageway.
In a further improvement, the congestion degree of the train delayed by a basic time granularities is calculated by the following formula (9):
in the formula (I), the compound is shown in the specification,
the passenger flow density index of the t 1min line;
Further improvement, the train delays the crowdedness degree theta under a basic time granularityCongestion aCalculated from equation (12):
in the formula, thetaCongestion iIs the congestion level at the ith basic time granularity.
The invention provides a Bayesian network-based dynamic risk analysis method for a high-speed rail contact network, which comprises the steps of firstly establishing a risk propagation chain model of insulator flashover of the high-speed rail contact network, carrying out dynamic mathematical modeling of characteristic quantity-dynamic probability on the risk chain model according to a physical mechanism generated by a fault, placing the dynamic mathematical modeling in a Bayesian network for calculation, and obtaining the risk occurrence probability by the model; determining the dynamic severity of risks in two dimensions of time and space by using the passenger flow intensity index, taking the basic time granularity and a station as basic units for obtaining real-time monitoring data, taking the propagation path and the propagation process of the risks as the principle of calculating the passenger flow intensity index, and calculating different severity of risk consequences under the condition that trains delay different time scales by layer; and combining the dynamic probability with the severity, and obtaining the dynamic risk level of the risk chain according to the risk assessment matrix. Corresponding suggestions and countermeasures are given for different levels of risk levels. Finally, the method is verified through a calculation example, so that the dynamic risk level of the high-speed rail contact network under the severe weather condition can be analyzed and evaluated, and a theoretical basis is provided for timely blocking of risks.
Drawings
FIG. 1 is a schematic view of a high speed railway contact net system;
FIG. 2 is a flow chart of a Bayesian network-based dynamic risk analysis method for a high-speed rail catenary;
FIG. 3 is a schematic view of a flashover risk propagation chain of an insulator of the contact network;
FIG. 4 is a diagram of a Bayesian network structure of a flashover risk propagation chain of a high-speed rail contact network;
FIG. 5 is a graph showing the relationship between the degree of contamination and the flashover probability;
FIG. 6 is a plot of leakage current maximum versus flashover probability;
FIG. 7 is a graph of the relationship between the short-circuit current and the duration of the short-circuit current and the probability of disconnection;
FIG. 8 is a graph of line intensity index extrapolation;
FIG. 9 is a schematic diagram of a forming high-speed rail circuit;
FIG. 10 shows the Netica probability simulation results, where (a) is the simulation result of initial condition 1; (b) initial condition 2 simulation results.
The steps illustrated in the flow charts of the figures may be performed in a computer system such as a set of computer-executable instructions. Although a logical order is shown in the flow diagrams, in some cases, the steps described may be performed in an order different than here.
Detailed Description
Since the method of the present invention is described as being implemented in a computer system, the computer system may be provided in a processor of a server or a client. For example, the methods described herein may be implemented as software executable with control logic that is executed by a CPU in a server. The functionality described herein may be implemented as a set of program instructions stored in a non-transitory tangible computer readable medium. When implemented in this manner, the computer program comprises a set of instructions which, when executed by a computer, cause the computer to perform a method capable of carrying out the functions described above. Programmable logic may be temporarily or permanently installed in a non-transitory tangible computer-readable medium, such as a read-only memory chip, computer memory, disk, or other storage medium. In addition to being implemented in software, the logic described herein may be embodied using a discrete component, an integrated circuit, programmable logic used in conjunction with a programmable logic device such as a Field Programmable Gate Array (FPGA) or microprocessor, or any other device including any combination thereof. All such implementations are within the scope of the present invention.
Example 1
The embodiment 1 of the invention provides a method for analyzing the dynamic risk of a high-speed rail catenary based on a Bayesian network, as shown in FIG. 2, the method comprises the following steps:
s1: establishing a risk propagation chain model of the high-speed rail contact network insulator flashover, solving dynamic probabilities corresponding to all characteristic quantities in the risk propagation chain model, and substituting all the dynamic probabilities into the Bayesian network to obtain risk occurrence probabilities;
in step S1, specifically, taking an event that the insulator is polluted and causes the high-speed rail to fail to operate normally as an example, a risk propagation chain as shown in fig. 3 is established, and the mechanism is as follows:
with the continuous increase of the operation time, more and more pollutants are accumulated on the surface area of the insulator which is always exposed in the surrounding air, for example, dust generated in wind and sand weather, waste gas and slag generated in factories and the like are gradually accumulated on the surface layer of the insulator, so that a pollution layer is formed. Generally, the filth layers are kept dry and have no electric conduction capability, but in wet weather, the filth becomes gradually wet, the electrolyte contained in the filth layers is ionized under certain conditions, so that the conductive filth layers appear on the surface layer of the insulator, the resistance of the insulator is slowly reduced, the electric conduction is increased, and leakage current is formed under the action of voltage. Because the surface dirt of the insulator is uneven, the density of leakage current is different, the leakage current generates heat to evaporate surface moisture, thereby forming a dry band, the resistance at the dry band is very large, the voltage of the dry band is almost equal to the voltage at two ends of the insulator, and when the field intensity at the dry band is larger than the critical field intensity of the air discharge along the surface of the insulator, partial discharge occurs. If localized arcing occurs on both sides of the belt and the temperature is relatively high, the area of the belt gradually increases. If the voltage across the dry band does not make the arc burn, the local arc of the insulator will gradually shrink until it is extinguished, and the leakage current of the system will also gradually decrease. However, when the humidity of the environment where the insulator is located is high due to fog, fur rain and the like, the conductance of the dirt layer is increased, the leakage current is increased, a local arc discharge phenomenon can be generated again, the arc can be slowly diffused along with the power line, and finally the anode and the cathode of the insulator can be broken down along the surface to generate a flashover phenomenon.
After the insulator flashover occurs, the contact net generates two-phase short circuit, and the maximum short-circuit current can reach about 10kA instantly. If the contact network current protection time setting value is short, the main circuit breaker can be caused to trip, large-area multipoint pollution flashover accidents are easy to happen due to the fact that the insulator is similar in dirt accumulation and damp conditions in a pollution flashover area, and automatic reclosure cannot be coincided if flashover happens continuously, so that delay of different time can be caused to a train; if the contact line current protection time setting value is longer, the temperature of the contact line rapidly rises and softens, the contact line is broken when the contact line exceeds the tolerance tension, and the power supply is interrupted, so that train delay and station passenger detention and crowding can be caused, the railway transportation safety can be influenced, and the whole transportation performance can be greatly influenced.
In step S1, a Bayesian Network (BN) is a directed acyclic graph with probability annotations, in which nodes represent random variables abstracted from actual problems; directed arcs between nodes reflect associations between variables.
Let the sample space of experiment E be S, and a be the event of E. B is1,B2,…,BnA set of events that is E, and satisfies: b is1,B2,…,BnAre not mutually compatible, B1∪B2∪…BnWhen S is expressed, a bayesian expression represented by formula (1) is given:
in the formula, P (B)i) Known as prior probability, P (B)i| A) is called the posterior probability, P (A | B)i) Referred to as likelihood.
A bayesian network diagram of a risk propagation chain established based on a bayesian network as shown in fig. 4, the risk probability of congestion of passenger flow caused by power supply interruption of a high-speed rail is evaluated according to a known node probability distribution.
The state variables indicated by the letters in fig. 4 and the values of the state variables are shown in table 1.
TABLE 1 State variable value-taking Table
And selecting the characteristic quantity pollution degree, the leakage current and the short circuit current in the risk transmission chain model as main monitoring characteristic quantities of the flashover probability of the contact network insulator based on the established risk transmission chain model. And then the accuracy of the risk occurrence probability is obviously improved.
In step S1, the method of obtaining the dynamic probability corresponding to each feature amount is as follows:
1) calculation of dynamic probability P (D | a):
p (D | a) is the flashover probability of an event that is determined on the condition of the degree of contamination.
The dirty area grades are divided according to three factors of a dirty source, meteorological conditions, equivalent salt deposit density and the like, and are shown in table 2.
TABLE 2 Classification basis of filth classes
For the sake of clarity, the pollution grades are respectively assigned as {0,1,2,3,4}, and the pollution degree is gradually increased from 0 grade to 4 grades, wherein 0 grade is a clean area and 4 grades are a severely polluted area.
As can be seen from the table, the pollution degree can be divided into five grades, the higher the pollution degree is, the higher the flashover probability is, the relation between the pollution degree and the flashover probability is shown in FIG. 5, wherein P in the graph is1、P2、P3、P4、P5The value of (b) can be adjusted accordingly based on expert experience and actual operating environment.
2) Calculation of dynamic probability P (D | B):
p (D | B) is the flashover probability obtained for the event conditioned on the leakage current.
The different development stages of the surface discharge of the insulator correspond to different flashover probabilities, and according to a pollution flashover mechanism, the development of the leakage current can be divided into the following four stages according to the different discharge characteristics:
first stage (safe zone): when the leakage current is small, the heat generated by the leakage current cannot evaporate the moisture in the dirt layer, and a dry band cannot be formed at the moment, so that the discharge phenomenon does not exist at the moment, the insulator is positioned in a safe area, and the flashover probability is 0;
second stage (forecast area): when the leakage current is gradually increased, joule heat is generated by the leakage current, the moisture at the position with higher leakage current density begins to evaporate due to the heat of the joule heat, so that the conductivity of a dirt layer is increased, and sporadic small sparks are generated when an insulator is in a dry band state;
stage three (precaution area): when the leakage current continues to increase, the conductivity is increased, orange continuous arc discharge with short duration gradually occurs in a dry zone, the discharge frequency of the arc is increased continuously, the energy is increased gradually, sporadic small spark discharge is weakened gradually, and at the moment, a large-penetrability arc is not formed;
fourth stage (hazard zone): the continuous increase in leakage current increases the arc length, which increases the likelihood of the arc penetrating the surface of the insulator, and flashover occurs at any time:
according to the discharge mechanism, the maximum value of leakage I is constructedmThe flashover probability P, the fitted curve is shown in FIG. 6.
As can be seen from the figure, when the leakage current is in the safe area, the probability of the insulator flashover is almost 0; when the current is in the prediction region, the flashover probability is low, but the leakage current can be continuously developed and increased, so that when the effective value of the leakage current is in the prediction region, attention needs to be paid to strengthen the monitoring of the leakage current; when the effective value of the leakage current is increased to the range of an early warning area, an alarm device needs to send out an alarm signal in time, positioning information is displayed at the same time, emergency handling work such as patrol and the like needs to be arranged immediately in production scheduling, and the hidden danger of flashover is eliminated in advance; when the effective value of the leakage current reaches the dangerous area, the probability of flashover is extremely high, the coordination work of dispatching, operation and maintenance of the train and the station is done, and the emergency plan of temporary train stop and delay is made, so that the phenomenon that the station detained crowds are blocked, trample accidents and the like after the train delay is prevented.
3) Calculation of dynamic probability P (E | C):
p (E | C) is the disconnection probability obtained by the event with the short-circuit current as a condition.
After the insulator of the contact network is in flashover, the short-circuit current which is generated instantaneously can reach a very large value, a large amount of joule heat can be generated along with the increase of the short-circuit duration time to cause the instantaneous temperature rise of the contact line, the line is softened, the mechanical strength is greatly reduced, and finally the contact line is broken under the tension.
The duration of the short-circuit current depends on the setting time t of the quick-break protection1Reaction time t of relay protection device2And trip time t of the circuit breaker body3And (4) summing. In actual operation, due to factors such as different models of various products, the duration time of the short-circuit current can be different, and with the increase of the duration time of the short-circuit current, when the short-circuit current value is larger, the temperature of the contact line can be rapidly increased, so that the probability of causing the disconnection fault is increased.
Cross section 100mm2The relation between the fusing current and the duration of the copper conductor is shown as the following formula (2):
in the formula I0Fusing current for the wire; t is tscThe short circuit current duration.
Since the conductor heats up more when a short circuit occurs, its resistance and specific heat capacity can no longer be considered as constants but rather as a function of temperature. The temperature rise degree of the conductor after the conductor short circuit is represented by using the thermal effect (the heat generation amount of unit resistance), and the thermal effect Q when the conductor short circuit reaches the maximum allowable temperature can be known according to the characteristics of conductor short circuit heating and a thermal balance equationkCan be calculated according to equation (3):
Qk＝S2(Ah-Aw) (3)
in the formula, S is the contact line sectional area; a. thewFor parameters relating to the working temperature and material of the contact line, AhIs a parameter related to the highest temperature of the contact line and the material of the contact line in short circuit, and has the unit of J/(omega m)4). Therefore, the temperature of the molten metal is controlled,for different types of contact lines, the three parameters are different, and in actual operation, specific numerical values can be obtained by looking up relevant data.
Because the time that the current passes through is very short when the current-carrying conducting wire is short-circuited, the emitted heat is not as long as being dissipated into the air, which can be considered as an adiabatic process, and all the heat generated by the short-circuit current is used for temperature rise, so the real-time heat effect Q of the conductor can be calculated according to the formula (4):
in the formula, I is the effective value of the current passing through the contact line; t is tscThe short circuit current duration.
When the contact network is short-circuited, the short-circuit impedances of different short-circuit points are different, so that the magnitude of the short-circuit current of each short-circuit point is greatly different, and the probability of wire disconnection caused by the temperature rise of a lead due to the short-circuit current can be expressed as a function of the short-circuit current and the duration time of the short-circuit current. The disconnection probability is the ratio of the real-time thermal effect of the short-circuit current to the thermal effect of the wire when the wire is fused, as shown in formula (5),
wherein I is the effective value of the short-circuit current, I0Is the fusing current of the conductor.
And after the three dynamic probabilities are obtained, substituting the three dynamic probabilities into the Bayesian network, and calculating to obtain the risk occurrence probability P. For this risk propagation chain, the risk occurrence probability P can be divided into several levels as shown in table 3:
TABLE 3 Risk occurrence probability rating criteria
| Grade | First stage | Second stage | Three-stage | Four stages | Five stages |
| Description of accidents | Tea table | Is low in | In | Height of | Super high |
| Interval probability | P＜0.01％ | 0.01％≤P<0.1％ | 0.1％≤P<1％ | 1％≤P<10％ | P≥10％ |
S2: calculating passenger flow intensity indexes according to risk propagation paths and propagation processes in the risk propagation chain model, and calculating the severity of risk consequences caused by train delay at different times;
in step S2, the severity of the consequences caused by the flashover of the contact network is represented by the product of the passenger flow intensity index and the train delay time, as shown in formula (6):
Yzcd＝K×t (6)
in the formula, Yzcd represents the severity, K represents the line passenger flow density index, and t represents the train delay time.
The passenger flow density index is an index which comprehensively reflects the passenger flow distribution condition and the congestion degree of the rail transit, and the value of the passenger flow density index is closely related to the severity and the congestion range of the passenger flow congestion. The severity of passenger flow congestion is evaluated by using a passenger flow density index, the traditional passenger flow density is considered, and the influence of continuous large passenger flow on station and line safety can be reflected.
In terms of space, after the train stops running, the propagation path of the congestion risk is a station-line-road network; the severity of the consequences of congestion increases over time as the time of train outage increases.
The higher the passenger flow density degree is, the more serious the station congestion degree is, and the correspondence between the product of the passenger flow density index and the delay time and the congestion severity degree is established according to actual experience as shown in table 4.
Table 4 congestion severity ranking
In step S2, the method for estimating the line passenger flow intensity index specifically includes:
the dynamic propagation process of the congestion risk is station key areas-stations-whole lines, so the congestion severity of the whole line depends on the congestion severity of each station, namely the congestion severity level of the whole line can be calculated on the basis of knowing the passenger flow density of each station. The passenger flow intensity index of each station can be calculated by using daily monitoring data, 1min is selected as a basic time granularity, and under the condition that the intensity index of 1min at the station is known, the passenger flow intensity index calculation method of the line specifically comprises the following calculation processes:
(1) the congestion degree of the line under the basic time granularity can be obtained according to the formula (7)
In the formula, alpha11、α12…α1iThe weights of the 1 st station, the 2 nd station and the i th station are respectively; mu.s1、μ2To muiPassenger flow density indexes of 1 st station, 2 th station and i th station respectively;
the degree of crowding of the station for 1min under the condition of normal running of the train is shown.
When the train runs in a power supply stopping accident, the influence of each station on the crowdedness degree of the whole line is changed, namely the weight is changed. When a train breaks down, the dynamic weight of each station on the line can be regarded as a value which is gradually decreased from a fault center and is recorded as alpha according to different fault positions and different influence degrees on the stations2iThen, the total weight α of each station is given according to equation (8):
α＝α1iα2i(8)。
(2) according to the specific situation of each station, reasonably setting the density threshold of each station, and taking the station proportion higher than the threshold in the line as the congestion range theta of the lineCongestion of the earthSpecifically, it is calculated according to formula (11):
in the formula phiareaThe passenger flow density value threshold value of each basic observation unit is given for each station; phijThe passenger flow density value of the jth basic observation unit of the station; mjThe number of the basic observation units exceeding the passenger flow density value threshold value is N, and the N is the total number of the basic observation units.
(3) By combining the congestion degree and the congestion range of the line, the passenger flow density index of the line within 1min can be calculated in the passenger flow density index calculation graph, as shown in fig. 8, and the severity of the congestion risk consequence of the whole line can be obtained by grading according to the congestion severity.
(4) When the train delay time is different, the passenger flow density index in the delay time period can be calculated according to the specific delay time and the line density index in the basic time granularity, so that the crowding degrees of different delay times are obtained, as shown in formula (9):
in the formula (I), the compound is shown in the specification,
the passenger flow density index of the t 1min line;
(5) Degree of congestion theta of train delayed by a basic time granularityCongestion aCalculated from equation (12):
in the formula, thetaCongestion iIs the congestion degree at the ith basic time granularity
Taking the passenger flow density index of which a is 5 as an example, a basic time granularity is 5min, monitoring and counting the passenger flow density index every 1min within 5min to obtain 5 time sequence data. Reasonably setting a passenger flow density index threshold value of the line for 1min according to the experience of management personnel, taking the ratio of the sum of density indexes higher than the threshold value within 5min to the total passenger flow density index as the congestion degree of the passenger flow within 5min,
in the formula (I), the compound is shown in the specification,
the passenger flow density index of the t 1min line;
a threshold value of the line 1min passenger flow intensity index;
the congestion level of the line is 5 lines and 1 min. According to the formula (12), the line is delayed by 5 and within the congestion range theta of 1minCrowding for 5min. By
And thetaCrowding for 5minThe passenger flow density index of the line for 5min can be obtained according to the density index calculation graph, and the congestion severity of the line for 5min later can be obtained according to the classification of risk consequences.
S3: the Risk assessment matrix Risk shown in the formula (10) is given based on the Risk occurrence probability obtained in step S1 and the severity obtained in step S2,
Risk＝f(p，c) (10)
p is the risk occurrence probability and c is the severity;
in step S3, the risk assessment matrix is shown in table 5:
TABLE 5 Risk assessment matrix
In the formula, from R1To R5The risk level is gradually increased, and can be defined from low to high as 1 to 5 according to the needs.
The risk caused by the insulator flashover of the contact network can be avoided by taking different measures for different grades of risks according to the table 6. In practical application, corresponding measures can be made according to different grades, so that risk propagation is blocked in time, and loss caused by risks is reduced.
TABLE 6 grading of flashover risk indicators for contact networks
S4: and evaluating the dynamic risk of the high-speed rail overhead line system according to the risk evaluation matrix.
The risk evaluation matrix provided by the invention can evaluate the possible risks in real time, so that corresponding suggestions are provided for blocking the risk propagation chain, and various safety problems are avoided.
Example 2
The evaluation was conducted by taking the forming high iron as an example, and the running route of the forming high iron is shown in fig. 9.
The Yu forming area is subtropical monsoon climate, and the area has the obvious characteristics of cloud fog, short sunshine time and higher probability of flashover of a train contact net. Based on the actual situation of the Yu-forming area, it is assumed that a certain train which is monitored to run in the North Longchang-Rongchang station respectively operates under the two conditions with different initial environmental conditions as shown in Table 7.
TABLE 7 Bayesian simulation initial condition evaluation
| Initial conditions | |
|
| Initial distribution of foul conditions (salt density) | ρESDD～N(0.07,0.1) | ρESDD～N(0.25,0.15) |
| Maximum leakage current monitoring magnitude | 124mA | 350mA |
| Contact line model | Pure copper contact line TB/T2810 | Pure copper contact line TB/T2810 |
| Contact line cross section | 100mm2 | 100mm2 |
| Short circuit current measurement | 4.6kA | 8kA |
According to relevant research and expert experience, three dynamic probability function relations are respectively assigned, and a simulation result is calculated by using Bayesian professional simulation software Netica, as shown in FIG. 10.
It can be seen from the figure that when the dynamic probability of each event changes due to the change of the external environment, the probability distribution of each event in the bayesian network also changes dynamically, and the probability of the final passenger congestion caused by the pollution flashover rises from 0.43% in (a) to 17.5% in (b), because the pollution level of the whole system is relatively low when the train runs in the area under the given condition 1, and therefore the whole risk probability is at a lower level. According to table 4, the simulation results in the two cases show that the probability of congestion at the station is at a medium level and at an extremely high level, respectively.
Determining the initial weight of each station on the Yu high-speed railway according to the daily traffic of passengers entering and leaving the station; when the train stops running in the region of Longchang North-Rongchang North, the dynamic weight of each station on the line is considered to decrease from the two stations of Longchang North and Rongchang North to the outside. If the train is 5min later, respectively assuming that the passenger flow density index of each station for 5min and the weight of each station are known according to the actual situation, and the specific value taking situation is shown in table 8.
Table 8 Yu high-speed railway station passenger flow intensity index and weight assignment
As can be seen from the above-mentioned estimated relationship,
assuming that all station intensity thresholds are 6, then θCongestion of the earth0.25. According to the following graph, the line passenger flow density index which can be caused by the power failure delay of the train between the Longchang north station and the Rongchang north station is calculated to be 3, and the congestion severity is moderate and severe. And combining the probabilities obtained by simulation under two operating environments, and obtaining that the overall risk level of the system is respectively at a medium level and a high level according to the risk assessment matrix. At this point, corresponding recommendations may be given by table 6 based on different risk levels and measures may be taken to block the spread of risk.
Claims (10)
1. A high-speed rail overhead line system dynamic risk analysis method based on a Bayesian network is characterized by comprising the following steps:
s1: establishing a risk propagation chain model of the high-speed rail contact network insulator flashover, solving dynamic probabilities corresponding to all characteristic quantities in the risk propagation chain model, and substituting all the dynamic probabilities into the Bayesian network to obtain risk occurrence probabilities;
s2: calculating passenger flow intensity indexes according to risk propagation paths and propagation processes in the risk propagation chain model, and calculating the severity of risk consequences caused by train delay at different times;
s3: the Risk assessment matrix Risk shown in the formula (10) is given based on the Risk occurrence probability obtained in step S1 and the severity obtained in step S2,
Risk＝f(p，c) (10)
p is the risk occurrence probability and c is the severity;
s4: and evaluating the dynamic risk of the high-speed rail overhead line system according to the risk evaluation matrix.
2. The Bayesian network-based high-speed rail catenary dynamic risk analysis method according to claim 1, wherein the characteristic quantities in the risk propagation chain model include pollution degree, leakage current and short-circuit current.
3. The bayesian-net-based high-speed rail catenary dynamic risk analysis method of claim 1, wherein the severity Yzcd is characterized by formula (6):
Yzcd＝K×t (6)
k represents the line passenger flow intensity index, and t is train delay time.
4. The method for analyzing the dynamic risk of the high-speed rail catenary based on the bayesian network as claimed in claim 1, wherein the risk assessment matrix of step S3 is as follows:
in the table, from R1To R5The risk level is gradually increased, and is defined as 1 to 5 levels from low to high as required.
5. The Bayesian net-based high-speed rail catenary dynamic risk analysis method as claimed in claim 2, wherein the dynamic probability comprises a disconnection probability P (E | C) calculated by equation (5):
i is the effective value of the short-circuit current passing through the contact line; i is0Fusing current for the wire; t is tscFor short-circuit current duration, S is contactArea of wire cross section; a. thewFor parameters relating to the working temperature and material of the contact line, AhIs the parameter related to the highest temperature of the contact line and the material of the contact line in short circuit.
6. The Bayesian network-based high-speed rail catenary dynamic risk analysis method as claimed in claim 1, wherein the passenger flow intensity index is derived from the congestion degree and the congestion range of the line.
7. The Bayesian net-based high-speed rail catenary dynamic risk analysis method of claim 6, wherein the degree of crowding at a basic time granularity is
Calculated according to equation (7):
in the formula, alpha11、α12...α1iThe weights of the 1 st station, the 2 nd station and the i th station are respectively; mu.s1、μ2To muiPassenger flow density indexes of 1 st station, 2 th station and i th station respectively;
the degree of crowding of the station for 1min under the condition of normal running of the train is shown.
8. The Bayesian network-based high-speed rail catenary dynamic risk analysis method of claim 6, wherein the crowding range θ at the basic time granularityCrowding for 1minCalculated from equation (11):
in the formula phiareaPassenger flow density value threshold of each basic observation unit given for each stationA value; phijThe passenger flow density value of the jth basic observation unit of the station; mjThe number of the basic observation units exceeding the passenger flow density value threshold value is N, and the N is the total number of the basic observation units.
9. The Bayesian network-based high-speed rail catenary dynamic risk analysis method according to claim 6, wherein the degree of congestion of a train delayed by a basic time granularity is calculated by equation (9):
in the formula (I), the compound is shown in the specification,
the passenger flow density index of the t 1min line;
a threshold value of the line 1min passenger flow intensity index;
is the congestion level of a basic time granularity of the line.
10. The Bayesian network-based high-speed rail catenary dynamic risk analysis method according to claim 9, wherein the degree of congestion θ at a basic time granularity of a train delayCongestion aCalculated from equation (12):
in the formula, thetaCongestion iIs the congestion level at the ith basic time granularity.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910165009.0A CN109948204B (en) | 2019-03-05 | 2019-03-05 | Bayesian network-based dynamic risk analysis method for high-speed rail overhead line system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910165009.0A CN109948204B (en) | 2019-03-05 | 2019-03-05 | Bayesian network-based dynamic risk analysis method for high-speed rail overhead line system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109948204A CN109948204A (en) | 2019-06-28 |
| CN109948204B true CN109948204B (en) | 2020-11-03 |
Family
ID=67008504
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910165009.0A Active CN109948204B (en) | 2019-03-05 | 2019-03-05 | Bayesian network-based dynamic risk analysis method for high-speed rail overhead line system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109948204B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111806306B (en) * | 2020-06-05 | 2021-11-19 | 中铁电气化勘测设计研究院有限公司 | Double-protection grounding wire method for direct-current traction power supply double-insulation contact network system |
| CN111652395A (en) * | 2020-06-12 | 2020-09-11 | 成都国铁电气设备有限公司 | Health assessment method for high-speed railway contact network equipment |
| CN112162528B (en) * | 2020-09-29 | 2022-02-15 | 广东工业大学 | Fault diagnosis method, device, equipment and storage medium of numerical control machine tool |
| CN112232415B (en) * | 2020-10-16 | 2021-07-20 | 中南大学 | Method, equipment and medium for identifying delay spread of high-speed rail full network station |
| CN112070325B (en) * | 2020-11-12 | 2021-02-26 | 北京交通大学 | Road network train optimization method, device, equipment and storage medium under abnormal event |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107145985A (en) * | 2017-05-09 | 2017-09-08 | 北京城建设计发展集团股份有限公司 | A kind of urban track traffic for passenger flow Regional Linking method for early warning |
| CN107403189A (en) * | 2017-06-30 | 2017-11-28 | 南京理工大学 | A kind of windage yaw discharge method for early warning based on Naive Bayes Classifier |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016123619A1 (en) * | 2015-01-30 | 2016-08-04 | New York University | System and method for electrophysiological monitoring |
| CN107368955A (en) * | 2017-06-29 | 2017-11-21 | 广西电网有限责任公司 | The implementation method of power grid risk, equipment Risk and operating risk integration linkage |
| CN107918705B (en) * | 2017-11-14 | 2020-09-18 | 山东电力工程咨询院有限公司 | Method for calculating installation necessity of overhead line arrester |
-
2019
- 2019-03-05 CN CN201910165009.0A patent/CN109948204B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107145985A (en) * | 2017-05-09 | 2017-09-08 | 北京城建设计发展集团股份有限公司 | A kind of urban track traffic for passenger flow Regional Linking method for early warning |
| CN107403189A (en) * | 2017-06-30 | 2017-11-28 | 南京理工大学 | A kind of windage yaw discharge method for early warning based on Naive Bayes Classifier |
Non-Patent Citations (2)
| Title |
|---|
| Effect of washing water flow rate and pollution level on leakage current of a fixed washed high voltage insulator;Mahmoud M. Daha等;《2016 Eighteenth International Middle East Power Systems Conference (MEPCON)》;20170202;全文 * |
| 基于贝叶斯网络的接触网运行可靠性评估;王佳培;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20170715;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109948204A (en) | 2019-06-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109948204B (en) | Bayesian network-based dynamic risk analysis method for high-speed rail overhead line system | |
| CN109784782B (en) | Subway power supply system dynamic risk analysis and evaluation method based on fuzzy inference | |
| Chen et al. | Reliability evaluations of railway power supplies by fault-tree analysis | |
| CN103207340B (en) | On-line transmission line lightning shielding failure trip early-warning method | |
| Velásquez et al. | Ruptures in overhead ground wire—Transmission line 220 kV | |
| US20210278464A1 (en) | Condition assessment method and device for an outdoor post-mounted vacuum switch | |
| CN108898258A (en) | The analysis method and system of cascading failure in power system risk under Lightning Disaster weather | |
| CN111680872B (en) | Power grid risk calculation method based on multi-source data fusion | |
| CN104103019A (en) | Operation risk assessment method and assessment system of power distribution network containing distributed power supply | |
| CN107194574A (en) | A kind of grid security risk assessment method based on load loss | |
| CN105467276A (en) | Line fault monitoring method and system | |
| CN104166940A (en) | Method and system for assessing power distribution network operation risk | |
| Ma et al. | A dynamic risk analysis method for high-speed railway catenary based on Bayesian network | |
| CN103218754A (en) | Risk test method and risk test device for power gird dispatching operation | |
| CN104112076A (en) | Fuzzy mathematics based operational risk assessment method and fuzzy mathematics based operational risk assessment system | |
| Feng et al. | Risk index system for catenary lines of high-speed railway considering the characteristics of time–space differences | |
| Gao et al. | A multilayer Bayesian network approach-based predictive probabilistic risk assessment for overhead contact lines under external weather conditions | |
| CN101860002A (en) | Computing method of preventive control measure for short-circuit current in electric power system | |
| Davoudi et al. | Reclosing of distribution systems for wildfire prevention | |
| CN109978345A (en) | A kind of bullet train trailer system combined failure dynamic risk analysis method based on characteristic quantity | |
| Ciapessoni et al. | A probabilistic risk-based security assessment tool allowing contingency forecasting | |
| Listyukhin et al. | Systematization and monitoring of quality parameters of overhead power transmission lines functioning | |
| RU2676889C1 (en) | Meteo monitoring system for the electric grid equipment damage probability predicting and the preventive and recovery works conducting evaluation | |
| Akintola et al. | A Critical Review of Distribution Substation System Reliability Evaluations | |
| Su et al. | Lightning Trip Warning Based on GA-BP Neural Network Technology |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |