Power transmission tower landslide risk prediction method and system

Abstract

The invention discloses a power transmission tower landslide risk prediction method. The method comprises the steps of obtaining historical landslide data information of a target power transmission line; obtaining a power transmission line landslide association rule based on correlation analysis; predicting to obtain a current landslide prediction state of the target line; acquiring current landslide data information of the target power transmission line; reasoning error data are obtained through calculation; obtaining a current association strength factor; constructing a Bayesian prediction network and predicting the landslide risk of the target power transmission line; and according to the landslide risk, the current landslide prediction state and the reasoning error data, calculating to obtain a power transmission tower landslide risk prediction result of the target power transmission line. The invention also discloses a system for realizing the power transmission tower landslide risk prediction method. According to the invention, the landslide risk prediction of the power transmission tower can be realized, and the reliability, the accuracy and the timeliness are higher.

Description

Technical Field

[0001] The invention belongs to the field of electrical automation, and in particular relates to a method and system for predicting landslide risks of a transmission tower. Background Art

[0002] With the development of economy and technology and the improvement of people's living standards, electricity has become an indispensable secondary energy source in people's production and life, bringing endless convenience to people's production and life. Therefore, ensuring the stable and reliable supply of electricity has become one of the most important tasks of the power system.

[0003] Transmission lines are an important part of the power system, and transmission towers are one of the most important parts of transmission lines. Transmission lines are located in complex and changeable environments, especially under the influence of heavy rainfall and engineering construction, which can easily cause soil erosion and slope sliding on the surface of the transmission line foundation, causing landslide disasters on the transmission foundation, thus greatly affecting the safety of transmission towers. Therefore, it is particularly important to monitor the landslide risk of transmission towers.

[0004] At present, the traditional landslide risk monitoring scheme for transmission towers still adopts the manual monitoring scheme, that is, the power system inspectors manually monitor the towers of the transmission lines to find the landslide risk of the transmission towers on site. However, this type of manual monitoring method is not only time-consuming and labor-intensive, but also has poor reliability, and often has a significant lag effect, that is, once the landslide risk of the transmission tower is discovered, it is already too late. Summary of the invention

[0005] One of the purposes of the present invention is to provide a method for predicting the risk of landslide of a transmission tower with high reliability, good accuracy and good timeliness.

[0006] A second object of the present invention is to provide a system for implementing the transmission tower landslide risk prediction method.

[0007] The method for predicting the risk of landslide of a transmission tower provided by the present invention comprises the following steps:

[0008] S1. Obtain historical landslide data information of the target transmission line;

[0009] S2. According to the historical landslide data information obtained in step S1, based on correlation analysis, the landslide association rules of the transmission line are obtained;

[0010] S3. According to the historical landslide data obtained in step S1 and the association rules obtained in step S2, the current landslide prediction state of the target line is predicted;

[0011] S4. Obtaining current landslide data information of the target transmission line;

[0012] S5. Calculate the inference error data according to the current landslide data information obtained in step S4 and the current landslide prediction state obtained in step S3;

[0013] S6. Obtaining a current association strength factor according to the current landslide data information obtained in step S4 and the association rule obtained in step S2;

[0014] S7. Combine the current correlation strength factor obtained in step S6 to construct a Bayesian prediction network and predict the landslide risk of the target transmission line;

[0015] S8. Based on the landslide risk obtained in step S7, the current landslide prediction state obtained in step S3 and the inference error data obtained in step S5, the transmission tower landslide risk prediction result of the target transmission line is calculated.

[0016] The step S1 of obtaining the historical landslide data information of the target transmission line specifically includes the following steps:

[0017] The historical landslide data information of the target transmission line includes the transmission line landslide geological hazard assessment status set, the transmission line basic geological data set, the transmission line engineering meteorological data set and the transmission line tower basic data set;

[0018] The transmission line landslide geological hazard assessment status set includes time, transmission line name, tower number, voltage level, tower type, foundation type, landslide risk assessment level and landslide data;

[0019] The basic geological data set for transmission lines includes geological type, soil moisture, slope, altitude, slope position, vegetation type and vegetation coverage;

[0020] The meteorological data set for transmission line projects includes cumulative rainfall, monthly average rainfall, daily rainfall, daily maximum temperature, average temperature, construction distance, and construction damage extent;

[0021] The transmission line tower foundation data set includes foundation form, foundation size, foundation depth, tower material type, force type, tower height, tower span and tower rotation angle.

[0022] The step S2, based on the historical landslide data information obtained in step S1, obtains the landslide association rules of the transmission line based on correlation analysis, which specifically includes the following steps:

[0023] The confidence level indicates the probability that a particular sample is distributed within a specified interval; the confidence level is calculated using the following formula:

[0024]

[0025] Where confidence(T b→T a ) is in the interval T b For type T a The confidence level of the sample; count(T a ∪T b ) is in the interval T b For type T a The number of samples; count(T b ) is the interval T b The total number of samples on ;

[0026] Support refers to the ratio of a specific sample to the total samples; support is calculated using the following formula:

[0027]

[0028] Where support(T b →T a ) is in the interval T b The type is T a The support of the sample; count(T) is the total number of samples;

[0029] The confidence threshold and support threshold are set, and according to the confidence and support of landslide risk in different intervals, the correlation between geological type, soil moisture, slope, altitude, slope position, vegetation type, vegetation coverage, cumulative rainfall, average monthly rainfall, daily rainfall, maximum temperature of the day, average temperature, construction distance, degree of construction damage, cumulative rainfall, average monthly rainfall, daily rainfall, maximum temperature of the day, average temperature, construction distance, degree of construction damage, foundation form, foundation size, foundation depth, tower material type, force type, tower height, tower span, tower angle and landslide risk is obtained, so as to obtain the landslide association rules of transmission lines.

[0030] Step S3 predicts the current landslide prediction state of the target line based on the historical landslide data obtained in step S1 and the association rules obtained in step S2, and specifically includes the following steps:

[0031] Initial prediction stage: The historical landslide data at the initial time T0 is used to obtain the association rules at the initial time, and then the prediction error at the initial time and the actual state at the initial time are combined to use the Bayesian network to predict the predicted state at the time T1; the prediction error at the initial time is 0;

[0032] Normal prediction stage: using the historical data at time T1, the association strength of the historical data at time T1 is obtained by using association rules, and then the prediction error is obtained based on the prediction result at time T0 for time T1 and the actual state at time T1. Finally, the current landslide prediction state of the target line at subsequent moments is predicted by combining the prediction error and the actual state at time T1.

[0033] Step S5, according to the current landslide data information obtained in step S4 and the current landslide prediction state obtained in step S3, calculates the inference error data, which specifically includes the following steps:

[0034] Calculate error data based on the current landslide data information obtained in step S4 and the current landslide prediction state obtained in step S3;

[0035] According to the set rules, the error data is judged and the inference error data is obtained:

[0036] If Δ<-a1, the inference error data value is -4;

[0037] If -a1≤Δ<-a2, the inference error data value is -3;

[0038] If -a2≤Δ<-a3, the inference error data value is -2;

[0039] If -a3≤Δ<-a4, the inference error data value is -1;

[0040] If Δ=0, the inference error data value is 0;

[0041] If 0<Δ≤b1, the inference error data takes the value of 1;

[0042] If b1<Δ≤b2, the inference error data value is 2;

[0043] If b2<Δ≤b3, the inference error data value is 3;

[0044] If Δ>b3, the inference error data value is 4;

[0045] Among them, a1, a2, a3, a4, b1, b2 and b3 are all set demarcation thresholds, and a1>a2>a3>a4>0, b3>b2>b1>0.

[0046] The step S6 of obtaining the current association strength factor according to the current landslide data information obtained in step S4 and the association rule obtained in step S2 specifically includes the following steps:

[0047] Combined with the association rules obtained in step S2, the position of the current landslide data information obtained in step S4 in the association rules is determined to obtain the corresponding current association strength factor; for example, in the association rules obtained in step S2, if the tower is in the range of 10m to 50m above sea level, the landslide association degree is 0.3; if the current landslide data information shows that the tower is 45m above sea level, then it can be known that it belongs to the range of 10m to 50m, and the current association strength from the altitude dimension is 0.3.

[0048] The current correlation strength factor obtained in step S6 is combined with the step S7 to construct a Bayesian prediction network and predict the landslide risk of the target transmission line, which specifically includes the following steps:

[0049] The Bayesian prediction network includes a directed acyclic graph and a conditional probability table; the directed acyclic graph represents the causal relationship between several random variables, and the directed acyclic graph is composed of nodes and directed arcs, and the directed arcs point from the parent node to the child node; the conditional probability table is the probability quantification value of the causal relationship between the variables of each node, indicating the probability distribution of the node variable under different value combinations of the corresponding parent node variable; the root node variable has a priori marginal probability;

[0050] Combining the prior edge probability and conditional probability table, we can get the table containing all nodes U={X1,X2,...,X n The joint probability distribution of} is expressed as

[0051]

[0052] Where P(U) is the joint probability, which means the probability of occurrence under all known conditions; Pa(X i ) is X i The set of all parent nodes; P(X i |Pa(X i )) is the known event Pa(X i ) occurs, P(X i )The probability of an event occurring;

[0053] When using the Bayesian prediction network for predictive reasoning, when new evidence appears, the information will be transmitted along the directed arc, causing the posterior probability of other related node variables to change. The change in the posterior probability reflects the change in the state of the node variable, thereby realizing the predictive reasoning of the state; if evidence V={V1,...,V n}, then the joint probability P(U) is updated to P(U,V), then for the node variable X i The probability of updating is

[0054]

[0055] Where P(X|V) is the probability of the P(X) event occurring when the known event P(V) occurs; P(V) represents the probability of all V events occurring;

[0056] The constructed Bayesian prediction network includes the target layer including the prediction results of landslide risk of transmission lines; the sub-target layer includes the disaster-prone environment, disaster-causing risk and disaster-bearing risk; the state quantity layer includes geological characteristics, topographic characteristics, geomorphic characteristics, meteorological factors, human activities, foundation characteristics, tower characteristics and structural characteristics; the multi-source information layer includes geological type, soil moisture, slope, altitude, slope position, vegetation type, vegetation coverage, cumulative rainfall, daily rainfall, construction distance, damage form, foundation form, foundation size, foundation depth, tower material type, commissioning life, tower height, tower spacing and tower rotation angle;

[0057] The constructed Bayesian prediction network has node descriptions as follows: the node state values ​​of the node variables of the target layer, subcomponent layer, and state quantity layer are all normal, attention, abnormal, severe, and extremely severe, among which normal means the normal state has the highest probability, attention means the attention state has the highest probability, abnormal means the abnormal state has the highest probability, severe means the severe state has the highest probability, and extremely severe means the extremely severe state has the highest probability;

[0058] The node status values ​​of the node variable of the association strength of the multi-source information layer are no association, weak association, medium association and strong association. No association indicates a probability of 0, weak association indicates a probability of 0% to 35%, medium association indicates a probability of 35% to 65%, and strong association indicates a probability of 65% to 100%.

[0059] Step S8 calculates the landslide risk prediction result of the transmission tower of the target transmission line according to the landslide risk obtained in step S7, the current landslide prediction state obtained in step S3 and the reasoning error data obtained in step S5, and specifically includes the following steps:

[0060] The landslide risk value obtained in step S7 ranges from 1 to 5, and a higher value indicates a greater landslide risk;

[0061] The value of the current landslide prediction state obtained in step S3 is 1 to 5, and a higher value indicates a greater landslide risk;

[0062] The value of the inference error data obtained in step S5 is -4 to 4;

[0063] Finally, the transmission tower landslide risk prediction result of the target transmission line is calculated using the following formula:

[0064]

[0065] Wherein H is the transmission tower landslide risk prediction result of the target transmission line; H1 is the landslide risk value obtained in step S7; H3 is the reasoning error data obtained in step S5; H2 is the current landslide prediction state value obtained in step S3;

[0066] The greater the predicted landslide risk result of the transmission tower of the target transmission line, the greater the landslide risk.

[0067] The present invention also provides a system for implementing the transmission tower landslide risk prediction method, comprising a history acquisition module, a correlation analysis module, a state prediction module, a real-time acquisition module, a reasoning error module, an association strength module, a network construction module and a risk prediction module; the history acquisition module, the correlation analysis module, the state prediction module, the real-time acquisition module, the reasoning error module, the association strength module, the network construction module and the risk prediction module are connected in series in sequence; the history acquisition module is used to acquire historical landslide data information of the target transmission line, and upload the data information to the correlation analysis module; the correlation analysis module is used to obtain the transmission line landslide association rules based on the received data information and the acquired historical landslide data information, and upload the data information to the state prediction module; the state prediction module is used to predict the current landslide prediction state of the target line based on the received data information, the acquired historical landslide data and the acquired association rules, and upload the data information to the real-time acquisition module ; The real-time acquisition module is used to obtain the current landslide data information of the target transmission line according to the received data information, and upload the data information to the reasoning error module; the reasoning error module is used to calculate the reasoning error data according to the received data information, the current landslide data information obtained and the current landslide prediction state obtained, and upload the data information to the association strength module; the association strength module is used to obtain the current association strength factor according to the received data information, the current landslide data information obtained and the association rules obtained, and upload the data information to the network construction module; the network construction module is used to construct a Bayesian prediction network according to the received data information and the current association strength factor obtained, and predict the landslide risk of the target transmission line, and upload the data information to the risk prediction module; the risk prediction module is used to calculate the transmission tower landslide risk prediction result of the target transmission line according to the received data information, the obtained landslide risk, the current landslide prediction state and the reasoning error data.

[0068] The method and system for predicting the landslide risk of a transmission tower provided by the present invention use a Bayesian network to predict the landslide risk of a transmission tower by analyzing and calculating historical data and combining the characteristics of current real-time data. Therefore, the present invention can not only realize the landslide risk prediction of the transmission tower, but also has higher reliability, better accuracy and better timeliness. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 The figure is a schematic diagram of the method flow of the present invention.

[0070] Figure 2 Schematic diagram of the Bayesian network model of the method of the present invention.

[0071] Figure 3 Schematic diagram of the functional modules of the system of the present invention. DETAILED DESCRIPTION

[0072] like Figure 1 The method flow chart of the method of the present invention is shown as follows: The method for predicting the risk of landslide of a transmission tower disclosed in the present invention comprises the following steps:

[0073] S1. Obtaining historical landslide data information of the target transmission line; specifically including the following steps:

[0074] The historical landslide data information of the target transmission line includes the transmission line landslide geological hazard assessment status set, the transmission line basic geological data set, the transmission line engineering meteorological data set and the transmission line tower basic data set;

[0075] The transmission line landslide geological hazard assessment status set includes time, transmission line name, tower number, voltage level, tower type, foundation type, landslide risk assessment level and landslide data; among them, the landslide risk assessment level is generally selected as 5 levels, and the larger the level, the greater the landslide risk;

[0076] The basic geological data set for transmission lines includes geological type, soil moisture, slope, altitude, slope position, vegetation type and vegetation coverage;

[0077] The meteorological data set for transmission line projects includes cumulative rainfall, monthly average rainfall, daily rainfall, daily maximum temperature, average temperature, construction distance, and construction damage extent;

[0078] The transmission line tower foundation data set includes foundation form, foundation size, foundation depth, tower material type, force type, tower height, tower span and tower rotation angle;

[0079] S2. According to the historical landslide data information obtained in step S1, based on correlation analysis, the landslide association rules of the transmission line are obtained; specifically, the steps include:

[0080] The confidence level indicates the probability that a particular sample is distributed within a specified interval; the confidence level is calculated using the following formula:

[0081]

[0082] Where confidence(T b →T a ) is in the interval T b For type T a The confidence level of the sample; count(T a ∪T b ) is in the interval T b For type Ta The number of samples; count(T b ) is the interval T b The total number of samples on ;

[0083] Support refers to the ratio of a specific sample to the total samples; support is calculated using the following formula:

[0084]

[0085] Where support(T b →T a ) is in the interval T b The type is T a The support of the sample; count(T) is the total number of samples;

[0086] Taking slope as an example, assuming that the total number of towers is 1000, the total number of towers with a slope of 0-10 degrees is 100, and 20 of them were found to have landslides by operation and maintenance. Then, for the slope dimension, the landslide support within the range of 0-10 degrees is 20 / 1000=2%, and the landslide confidence within the range of 0-10 degrees is 20 / 100=20%. In this way, the landslide association rules within the range of 0-10 degrees are formed.

[0087] The confidence threshold and support threshold are set. According to the confidence and support of landslide risk in different intervals, the correlation between geological type, soil moisture, slope, altitude, slope position, vegetation type, vegetation coverage, cumulative rainfall, average monthly rainfall, daily rainfall, maximum temperature of the day, average temperature, construction distance, degree of construction damage, cumulative rainfall, average monthly rainfall, daily rainfall, maximum temperature of the day, average temperature, construction distance, degree of construction damage, foundation form, foundation size, foundation depth, tower material type, force type, tower height, tower spacing, tower angle and landslide risk is obtained, so as to obtain the landslide association rules of transmission lines;

[0088] S3. According to the historical landslide data obtained in step S1 and the association rules obtained in step S2, the current landslide prediction state of the target line is predicted; specifically, the following steps are included:

[0089] Initial prediction stage: Since there is no prediction data in the initial Bayesian network training data set, the current status at time T0 is used as the predicted status at time T0; the historical landslide data at the initial time T0 is used to obtain the association rules at the initial time, and then the prediction error at the initial time and the actual status at the initial time are combined to use the Bayesian network to predict the predicted status at time T1; the prediction error at the initial time is 0;

[0090] Normal prediction stage: using the historical data at time T1, using the association rules to obtain the association strength of the historical data at time T1, and then obtaining the prediction error based on the prediction result at time T0 for time T1 and the actual state at time T1, and finally combining the prediction error with the actual state at time T1 to predict the current landslide prediction state of the target line at subsequent times;

[0091] S4. Obtain the current landslide data information of the target transmission line; in specific implementation, the data content obtained is the same as the data content obtained in step S1;

[0092] S5. Calculate the inference error data according to the current landslide data information obtained in step S4 and the current landslide prediction state obtained in step S3; specifically comprising the following steps:

[0093] Calculate error data based on the current landslide data information obtained in step S4 and the current landslide prediction state obtained in step S3;

[0094] According to the set rules, the error data is judged and the inference error data is obtained:

[0095] If Δ<-a1, the inference error data value is -4;

[0096] If -a1≤Δ<-a2, the inference error data value is -3;

[0097] If -a2≤Δ<-a3, the inference error data value is -2;

[0098] If -a3≤Δ<-a4, the inference error data value is -1;

[0099] If Δ=0, the inference error data value is 0;

[0100] If 0<Δ≤b1, the inference error data takes the value of 1;

[0101] If b1<Δ≤b2, the inference error data value is 2;

[0102] If b2<Δ≤b3, the inference error data value is 3;

[0103] If Δ>b3, the inference error data value is 4;

[0104] Among them, a1, a2, a3, a4, b1, b2 and b3 are all set demarcation thresholds, and a1>a2>a3>a4>0, b3>b2>b1>0;

[0105] S6. Obtaining a current association strength factor according to the current landslide data information obtained in step S4 and the association rule obtained in step S2; specifically comprising the following steps:

[0106] Combined with the association rule obtained in step S2, the position of the current landslide data information obtained in step S4 in the association rule is determined, so as to obtain the corresponding current association strength factor; for example, in the association rule obtained in step S2, if the tower is in the range of 10m to 50m above sea level, the landslide association degree is 0.3; if the current landslide data information shows that the tower is 45m above sea level, it can be known that it belongs to the range of 10m to 50m, and the current association strength from the altitude dimension is 0.3;

[0107] S7. Combining the current correlation strength factor obtained in step S6, constructing a Bayesian prediction network, and predicting the landslide risk of the target transmission line; specifically comprising the following steps:

[0108] The Bayesian prediction network includes a directed acyclic graph and a conditional probability table; the directed acyclic graph represents the causal relationship between several random variables, and the directed acyclic graph is composed of nodes and directed arcs, and the directed arcs point from the parent node to the child node; the conditional probability table is the probability quantification value of the causal relationship between the variables of each node, indicating the probability distribution of the node variable under different value combinations of the corresponding parent node variable; the root node variable has a priori marginal probability;

[0109] Combining the prior edge probability and conditional probability table, we can get the table containing all nodes U={X1,X2,...,X n The joint probability distribution of} is expressed as

[0110]

[0111] Where P(U) is the joint probability, which means the probability of occurrence under all known conditions; Pa(X i ) is X i The set of all parent nodes; P(X i |Pa(X i )) is the known event Pa(X i ) occurs, P(X i )The probability of an event occurring;

[0112] When using the Bayesian prediction network for predictive reasoning, when new evidence appears, the information will be transmitted along the directed arc, causing the posterior probability of other related node variables to change. The change in the posterior probability reflects the change in the state of the node variable, thereby realizing the predictive reasoning of the state; if evidence V={V1,...,V n}, then the joint probability P(U) is updated to P(U,V), then for the node variable X i The probability of updating is

[0113]

[0114] Where P(X|V) is the probability of the P(X) event occurring when the known event P(V) occurs; P(V) represents the probability of all V events occurring;

[0115] The constructed Bayesian prediction network is as follows: Figure 2 As shown in the figure, the target layer includes the landslide risk prediction results of the transmission line; the sub-target layer includes the disaster-prone environment, disaster-causing risk and disaster-bearing risk; the state quantity layer includes geological characteristics, topographic characteristics, geomorphic characteristics, meteorological factors, human activities, foundation characteristics, tower characteristics and structural characteristics; the multi-source information layer includes geological type, soil moisture, slope, altitude, slope position, vegetation type, vegetation coverage, cumulative rainfall, daily rainfall, construction distance, damage form, foundation form, foundation size, foundation depth, tower material type, commissioning life, tower height, tower span and tower angle;

[0116] The constructed Bayesian prediction network has node descriptions as follows: the node state values ​​of the node variables of the target layer, subcomponent layer, and state quantity layer are all normal, attention, abnormal, severe, and extremely severe, among which normal means the normal state has the highest probability, attention means the attention state has the highest probability, abnormal means the abnormal state has the highest probability, severe means the severe state has the highest probability, and extremely severe means the extremely severe state has the highest probability;

[0117] The node status values ​​of the node variable of the association strength of the multi-source information layer are no association, weak association, medium association and strong association. No association represents a probability of 0, weak association represents a probability of 0% to 35%, medium association represents a probability of 35% to 65%, and strong association represents a probability of 65% to 100%.

[0118] For example, the multi-source information layer includes geological type, soil moisture, slope, altitude, slope position, vegetation type, vegetation coverage, cumulative rainfall, daily rainfall, construction distance, damage form, foundation form, foundation size, foundation depth, tower material type, operation life, tower height, tower spacing, tower angle, etc. The association degree matrix can be obtained through association rules. Assuming it is [0.2, 0.3, 0.2, 0.1, 0.4, 0.2, 0.1, 0.3, 0.6, 0.1, 0.3, 0.2, 0.3, 0.4, 0.5, 0.4, 0.2, 0.3, 0.7], then combining the Bayesian network structure and the statistical conditional probability, the geological The overall risk level of the state quantity layer including characteristics, terrain characteristics, geomorphic characteristics, meteorological factors, human activities, basic characteristics, tower characteristics and structural characteristics; for example, the correlation strength of "accumulated rainfall" and "rainfall on the day" is [0.3, 0.6]. Through historical statistics, it is found that the correlation strength of the prior probability "accumulated rainfall" is 0.3, and there is a 60% prior probability that its "meteorological factor" level state quantity layer is a level II risk. Through historical statistics, it is found that the correlation strength of the prior probability "rainfall on the day" is 0.6, and there is an 80% prior probability that its "meteorological factor" level state quantity layer is a level V risk. Therefore, the highest risk level is taken, and the overall "meteorological factor" level state quantity layer is judged to be a level V risk. Then, a similar method is used to infer the risk level of the sub-target layer through the state quantity layer. Finally, the risk level of the sub-target layer is used to infer the overall landslide risk level of the transmission line;

[0119] S8. According to the landslide risk obtained in step S7, the current landslide prediction state obtained in step S3 and the reasoning error data obtained in step S5, the transmission tower landslide risk prediction result of the target transmission line is calculated; specifically comprising the following steps:

[0120] The landslide risk value obtained in step S7 ranges from 1 to 5, and a higher value indicates a greater landslide risk;

[0121] The value of the current landslide prediction state obtained in step S3 is 1 to 5, and a higher value indicates a greater landslide risk;

[0122] The value of the inference error data obtained in step S5 is -4 to 4;

[0123] Finally, the transmission tower landslide risk prediction result of the target transmission line is calculated using the following formula:

[0124]

[0125] Wherein H is the transmission tower landslide risk prediction result of the target transmission line; H1 is the landslide risk value obtained in step S7; H3 is the reasoning error data obtained in step S5; H2 is the current landslide prediction state value obtained in step S3;

[0126] The greater the predicted landslide risk result of the transmission tower of the target transmission line, the greater the landslide risk.

[0127] like Figure 3 The functional module schematic diagram of the system of the present invention is shown as follows: the system disclosed in the present invention for realizing the transmission tower landslide risk prediction method comprises a history acquisition module, a correlation analysis module, a state prediction module, a real-time acquisition module, a reasoning error module, an association strength module, a network construction module and a risk prediction module; the history acquisition module, the correlation analysis module, the state prediction module, the real-time acquisition module, the reasoning error module, the association strength module, the network construction module and the risk prediction module are connected in series in sequence; the history acquisition module is used to acquire the historical landslide data information of the target transmission line, and upload the data information to the correlation analysis module; the correlation analysis module is used to obtain the transmission line landslide association rules based on the received data information and the acquired historical landslide data information, and upload the data information to the state prediction module; the state prediction module is used to predict the current landslide prediction state of the target line based on the received data information, the acquired historical landslide data and the acquired association rules, and upload the data information to the state prediction module Upload real-time acquisition module; the real-time acquisition module is used to obtain the current landslide data information of the target transmission line according to the received data information, and upload the data information to the reasoning error module; the reasoning error module is used to calculate the reasoning error data according to the received data information, the current landslide data information obtained and the current landslide prediction state obtained, and upload the data information to the association strength module; the association strength module is used to obtain the current association strength factor according to the received data information, the current landslide data information obtained and the association rules obtained, and upload the data information to the network construction module; the network construction module is used to construct a Bayesian prediction network according to the received data information and the current association strength factor obtained, and predict the landslide risk of the target transmission line, and upload the data information to the risk prediction module; the risk prediction module is used to calculate the transmission tower landslide risk prediction result of the target transmission line according to the received data information, the obtained landslide risk, the current landslide prediction state and the reasoning error data.

Claims

1. A method for predicting the risk of landslide of a transmission tower, comprising the following steps: S1. Obtain historical landslide data information of the target transmission line; S2. According to the historical landslide data information obtained in step S1, based on correlation analysis, the landslide association rules of the transmission line are obtained; S3. According to the historical landslide data obtained in step S1 and the association rules obtained in step S2, the current landslide prediction state of the target line is predicted; S4. Obtaining current landslide data information of the target transmission line; S5. Calculate the inference error data according to the current landslide data information obtained in step S4 and the current landslide prediction state obtained in step S3; S6. Obtaining a current association strength factor according to the current landslide data information obtained in step S4 and the association rule obtained in step S2; S7. Combine the current correlation strength factor obtained in step S6 to construct a Bayesian prediction network and predict the landslide risk of the target transmission line; S8. Based on the landslide risk obtained in step S7, the current landslide prediction state obtained in step S3 and the inference error data obtained in step S5, the transmission tower landslide risk prediction result of the target transmission line is calculated.

2. The method for predicting the risk of landslide of a transmission tower according to claim 1, characterized in that The step S1 of obtaining the historical landslide data information of the target transmission line specifically includes the following steps: The historical landslide data information of the target transmission line includes the transmission line landslide geological hazard assessment status set, the transmission line basic geological data set, the transmission line engineering meteorological data set and the transmission line tower basic data set; The transmission line landslide geological hazard assessment status set includes time, transmission line name, tower number, voltage level, tower type, foundation type, landslide risk assessment level and landslide data; The basic geological data set for transmission lines includes geological type, soil moisture, slope, altitude, slope position, vegetation type and vegetation coverage; The meteorological data set for transmission line projects includes cumulative rainfall, monthly average rainfall, daily rainfall, daily maximum temperature, average temperature, construction distance, and construction damage extent; The transmission line tower foundation data set includes foundation form, foundation size, foundation depth, tower material type, force type, tower height, tower span and tower rotation angle.

3. The method for predicting the risk of landslide of a transmission tower according to claim 2, characterized in that The step S2, based on the historical landslide data information obtained in step S1, obtains the landslide association rules of the transmission line based on correlation analysis, which specifically includes the following steps: The confidence level indicates the probability that a particular sample is distributed within a specified interval; the confidence level is calculated using the following formula: Where confidence(T b →T a ) is in the interval T b For type T a The confidence level of the sample; count(T a ∪T b ) is in the interval T b For type T a The number of samples; count(T b ) is the interval T b The total number of samples on ; Support refers to the ratio of a specific sample to the total samples; support is calculated using the following formula: Where support(T b →T a ) is in the interval T b The type is T a The support of the sample; count(T) is the total number of samples; The confidence threshold and support threshold are set, and according to the confidence and support of landslide risk in different intervals, the correlation between geological type, soil moisture, slope, altitude, slope position, vegetation type, vegetation coverage, cumulative rainfall, average monthly rainfall, daily rainfall, maximum temperature of the day, average temperature, construction distance, degree of construction damage, cumulative rainfall, average monthly rainfall, daily rainfall, maximum temperature of the day, average temperature, construction distance, degree of construction damage, foundation form, foundation size, foundation depth, tower material type, force type, tower height, tower span, tower angle and landslide risk is obtained, so as to obtain the landslide association rules of transmission lines.

4. The method for predicting the risk of landslide of a transmission tower according to claim 3, characterized in that Step S3 predicts the current landslide prediction state of the target line based on the historical landslide data obtained in step S1 and the association rules obtained in step S2, and specifically includes the following steps: Initial prediction stage: The historical landslide data at the initial time T0 is used to obtain the association rules at the initial time, and then the prediction error at the initial time and the actual state at the initial time are combined to use the Bayesian network to predict the predicted state at the time T1; The prediction error at the initial moment is 0; Normal prediction stage: using the historical data at time T1, the association strength of the historical data at time T1 is obtained by using association rules, and then the prediction error is obtained based on the prediction result at time T0 for time T1 and the actual state at time T1. Finally, the current landslide prediction state of the target line at subsequent moments is predicted by combining the prediction error and the actual state at time T1.

5. The method for predicting the risk of landslide of a transmission tower according to claim 4, characterized in that Step S5, according to the current landslide data information obtained in step S4 and the current landslide prediction state obtained in step S3, calculates the inference error data, which specifically includes the following steps: Calculate error data based on the current landslide data information obtained in step S4 and the current landslide prediction state obtained in step S3; According to the set rules, the error data is judged and the inference error data is obtained: If Δ<-a1, the inference error data value is -4; If -a1≤Δ<-a2, the inference error data value is -3; If -a2≤Δ<-a3, the inference error data value is -2; If -a3≤Δ<-a4, the inference error data value is -1; If Δ=0, the inference error data value is 0; If 0<Δ≤b1, the inference error data takes the value of 1; If b1<Δ≤b2, the inference error data value is 2; If b2<Δ≤b3, the inference error data value is 3; If Δ>b3, the inference error data value is 4; Among them, a1, a2, a3, a4, b1, b2 and b3 are all set demarcation thresholds, and a1>a2>a3>a4>0, b3>b2>b1>0.

6. The method for predicting the risk of landslide of a transmission tower according to claim 5, characterized in that The step S6 of obtaining the current association strength factor according to the current landslide data information obtained in step S4 and the association rule obtained in step S2 specifically includes the following steps: Combined with the association rule obtained in step S2, the position of the current landslide data information obtained in step S4 in the association rule is determined, so as to obtain the corresponding current association strength factor.

7. The method for predicting the risk of landslide of a transmission tower according to claim 6, characterized in that The current correlation strength factor obtained in step S6 is combined with the step S7 to construct a Bayesian prediction network and predict the landslide risk of the target transmission line, which specifically includes the following steps: The Bayesian prediction network includes a directed acyclic graph and a conditional probability table; the directed acyclic graph represents the causal relationship between several random variables, and the directed acyclic graph is composed of nodes and directed arcs, and the directed arcs point from the parent node to the child node; the conditional probability table is the probability quantification value of the causal relationship between the variables of each node, indicating the probability distribution of the node variable under different value combinations of the corresponding parent node variable; the root node variable has a priori marginal probability; Combining the prior edge probability and conditional probability table, we can get the table containing all nodes U={X1,X2,...,X n The joint probability distribution of} is expressed as Where P(U) is the joint probability, which means the probability of occurrence under all known conditions; Pa(X i ) is X i The set of all parent nodes; P(X i |Pa(X i )) is the known event Pa(X i ) occurs, P(X i )The probability of an event occurring; When using the Bayesian prediction network for predictive reasoning, when new evidence appears, the information will be transmitted along the directed arc, causing the posterior probability of other related node variables to change. The change in the posterior probability reflects the change in the state of the node variable, thereby realizing the predictive reasoning of the state; if evidence V={V1,...,V n }, then the joint probability P(U) is updated to P(U,V), then for the node variable X i The probability of updating is Where P(X|V) is the probability of the P(X) event occurring when the known event P(V) occurs; P(V) represents the probability of all V events occurring; The Bayesian prediction network constructed includes the target layer including the landslide risk prediction results of the transmission line; the sub-target layer includes the disaster-prone environment, disaster-causing risk and disaster-bearing risk; the state quantity layer includes geological characteristics, terrain characteristics, geomorphic characteristics, meteorological factors, human activities, foundation characteristics, tower characteristics and structural characteristics; The multi-source information layer includes geological type, soil moisture, slope, altitude, slope position, vegetation type, vegetation coverage, cumulative rainfall, daily rainfall, construction distance, damage form, foundation form, foundation size, foundation depth, tower material type, years of operation, tower height, tower spacing and tower rotation angle; The constructed Bayesian prediction network has node descriptions as follows: the node state values ​​of the node variables of the target layer, subcomponent layer, and state quantity layer are all normal, attention, abnormal, severe, and extremely severe, among which normal means the normal state has the highest probability, attention means the attention state has the highest probability, abnormal means the abnormal state has the highest probability, severe means the severe state has the highest probability, and extremely severe means the extremely severe state has the highest probability; The node status values ​​of the node variable of the association strength of the multi-source information layer are no association, weak association, medium association and strong association. No association indicates a probability of 0, weak association indicates a probability of 0% to 35%, medium association indicates a probability of 35% to 65%, and strong association indicates a probability of 65% to 100%.

8. The method for predicting the risk of landslide of a transmission tower according to claim 7, characterized in that Step S8 calculates the landslide risk prediction result of the transmission tower of the target transmission line according to the landslide risk obtained in step S7, the current landslide prediction state obtained in step S3 and the reasoning error data obtained in step S5, and specifically includes the following steps: The landslide risk value obtained in step S7 ranges from 1 to 5, and a higher value indicates a greater landslide risk; The value of the current landslide prediction state obtained in step S3 is 1 to 5, and a higher value indicates a greater landslide risk; The value of the inference error data obtained in step S5 is -4 to 4; Finally, the transmission tower landslide risk prediction result of the target transmission line is calculated using the following formula: Wherein H is the transmission tower landslide risk prediction result of the target transmission line; H1 is the landslide risk value obtained in step S7; H3 is the reasoning error data obtained in step S5; H2 is the current landslide prediction state value obtained in step S3; The greater the predicted landslide risk result of the transmission tower of the target transmission line, the greater the landslide risk.

9. A system for implementing the method for predicting the risk of landslide of a transmission tower according to any one of claims 1 to 8, characterized in that It includes a history acquisition module, a correlation analysis module, a state prediction module, a real-time acquisition module, a reasoning error module, an association strength module, a network construction module and a risk prediction module; the history acquisition module, the correlation analysis module, the state prediction module, the real-time acquisition module, the reasoning error module, the association strength module, the network construction module and the risk prediction module are connected in series in sequence; the history acquisition module is used to acquire the historical landslide data information of the target transmission line, and upload the data information to the correlation analysis module; the correlation analysis module is used to obtain the landslide association rules of the transmission line based on the received data information and the acquired historical landslide data information, and upload the data information to the state prediction module; the state prediction module is used to predict the current landslide prediction state of the target line based on the received data information, the acquired historical landslide data and the acquired association rules, and upload the data information to the real-time acquisition module; The real-time acquisition module is used to obtain the current landslide data information of the target transmission line according to the received data information, and upload the data information to the reasoning error module; The reasoning error module is used to calculate the reasoning error data according to the received data information, the current landslide data information obtained and the current landslide prediction state obtained, and upload the data information to the association strength module; the association strength module is used to obtain the current association strength factor according to the received data information, the current landslide data information obtained and the association rules obtained, and upload the data information to the network construction module; The network construction module is used to construct a Bayesian prediction network based on the received data information and the current correlation strength factor, and predict the landslide risk of the target transmission line, and upload the data information to the risk prediction module; The risk prediction module is used to calculate the transmission tower landslide risk prediction result of the target transmission line according to the received data information, the obtained landslide risk, the current landslide prediction state and the reasoning error data.

Images