Working voltage-based high-altitude operation personnel safety prompting method and system

Abstract

The invention relates to the technical field of high-voltage line maintenance, and discloses a high-altitude operation personnel safety prompting method based on working voltage, which comprises the following steps: screening an accident distance between a historical high-voltage line and historical operation personnel and a preliminary factor of the accident distance, and splicing the working voltage and the preliminary factor, analyzing the grey correlation degree between the splicing factors and the accident distance, and selecting influence factors of the accident distance from the splicing factors; constructing a distance analysis model between the influence factor and the accident distance, and outputting a safety distance corresponding to the influence index; generating an initial factor and a safety value between the historical high-voltage line and the historical operating personnel, analyzing a factor-safety correlation degree between the initial factor and the safety value, and extracting a safety factor of the safety value from the initial factor; and constructing a safety early warning model between the safety factor and the safety value, and outputting a safety level corresponding to the safety index through the safety early warning model. According to the invention, safety early warning can be carried out on the premise of the dynamic safety distance.

Description

Technical Field

[0001] The invention relates to a method and system for providing safety reminders to high-altitude workers based on working voltage, and belongs to the technical field of high-voltage line maintenance. Background Art

[0002] With the development of society, the demand for electricity continues to grow, the scale of the power grid continues to expand, and high-voltage lines, as a key part of power transmission, have an increasingly wide coverage. This large-scale expansion of the power system has made the safety of high-voltage lines increasingly important. Once a high-voltage line fails, such as a short circuit or a broken line, it will instantly release huge amounts of energy. For example, the short-circuit current may reach thousands of amperes, generating great heat and impact force, posing a serious threat to the lives of workers.

[0003] At present, high-voltage lines carry high voltage and large current. Taking the common 110kV, 220kV or even higher voltage level high-voltage lines as an example, there is a strong electric field around them. When the human body is close to the high-voltage line, arc discharge may occur between the human body and the high-voltage line. For example, when the distance between the human body and the high-voltage line is less than the safe distance, the air may be broken down and conductive. For 110kV high-voltage lines, the safe distance is generally required to be more than 1.5 meters. This makes the safe distance between the high-voltage line and the maintenance personnel change dynamically under different high-voltage line environments. The existing technology sets a fixed distance between the high-voltage line and the maintenance personnel based on a single working voltage or a single working current. The set distance is fixed and unchanged, and the basis for setting the distance is relatively single. Secondly, whether the high-voltage line poses a threat to the maintenance personnel is not only related to whether the safe distance is exceeded. In other words, even if the maintenance personnel cross the safe distance, it does not mean that the maintenance personnel are in a dangerous state. This requires an early warning of the safety of the maintenance personnel.

[0004] Therefore, there is an urgent need for a solution that can provide safety warnings under the premise of a dynamic safety distance. Summary of the invention

[0005] The present invention provides a method and system for providing safety reminders to high-altitude workers based on working voltage, the main purpose of which is to provide safety warnings under the premise of a dynamic safety distance.

[0006] To achieve the above object, the present invention provides a method for providing safety reminders to high-altitude workers based on working voltage, comprising:

[0007] Collect historical accident logs and working voltages of historical high-voltage lines, select accident distances between the historical high-voltage lines and historical operators and preliminary factors of the accident distance from the historical accident logs, splice the working voltage and the preliminary factors to obtain splicing factors, analyze the grey correlation between the splicing factors and the accident distance, and select influencing factors of the accident distance from the splicing factors according to the grey correlation;

[0008] Constructing a distance analysis model between the influencing factors and the accident distance, collecting the influencing index between the current high-voltage line and the current operator, inputting the influencing index into the distance analysis model, and outputting the safe distance corresponding to the influencing index through the distance analysis model;

[0009] Collecting historical safety logs of the historical high-voltage line, generating initial factors and safety values ​​between the historical high-voltage line and historical operators using the safety distance and the historical safety logs, analyzing the factor-safety correlation between the initial factors and the safety values, and extracting the safety factors of the safety values ​​from the initial factors using the factor-safety correlation;

[0010] Constructing a safety warning model between the safety factor and the safety value, collecting the safety index between the current high-voltage line and the current operator, inputting the safety index into the safety warning model, and outputting the safety level corresponding to the safety index through the safety warning model;

[0011] The safety level is notified to the current operator to complete the safety prompt for the current operator and obtain the safety prompt result for the current operator.

[0012] Optionally, the screening of the accident distance between the historical high-voltage line and the historical operator and the preliminary factors of the accident distance from the historical accident log includes:

[0013] Querying the accident address between the historical high-voltage line and the historical operator from the historical accident log;

[0014] Performing address clustering on the accident address to obtain a clustered address;

[0015] identifying an address center of the clustered addresses;

[0016] Calculating the initial distance between the location of the historical high-voltage line and the center of the address;

[0017] Performing distance preprocessing on the initial distance to obtain the accident distance;

[0018] Extracting accident keywords and accident values ​​of the accident keywords from the historical accident log;

[0019] Divide the high-voltage line internal factors, personnel internal factors and environmental external factors between the historical high-voltage line and the historical operating personnel;

[0020] Searching for target keywords belonging to the internal cause of the high-voltage line, the internal cause of the personnel, and the external cause of the environment from the accident keywords;

[0021] Acquire a target value corresponding to the target keyword from the accident value;

[0022] Arranging the target values ​​belonging to the same target keyword among the target values ​​to obtain a target sequence;

[0023] Performing target preprocessing on the target sequence to obtain a preprocessed sequence;

[0024] The target keyword and the preprocessing sequence are used to generate preliminary factors of the accident distance.

[0025] Optionally, the analyzing the grey correlation between the splicing factor and the accident distance includes:

[0026] Obtaining the working voltage and preliminary factors in the splicing factors;

[0027] Performing voltage preprocessing on the working voltage to obtain a preprocessing voltage;

[0028] splicing the preprocessing voltage and the preliminary factor to obtain a splicing result;

[0029] The grey correlation coefficient between the splicing result and the accident distance is calculated using the following formula:

[0030]

[0031] Among them, i (k) represents the grey relational coefficient, x i (k) represents the kth value of the data of the ith category in the splicing result, x0(k) represents the kth cluster value of the accident distance, x j (k) means division by x i (k) is the kth value of the data of the jth category outside the range, ρ represents a constant, min k The symbol for taking the minimum value, max k The symbol for taking the maximum value;

[0032] According to the grey correlation coefficient, the grey correlation degree between the splicing result and the accident distance is calculated using the following formula:

[0033]

[0034] Among them, r 0i represents the grey correlation between the ith target keyword and the accident distance, ζ i (k) represents the grey relational coefficient, and m represents the number of k.

[0035] Optionally, constructing a distance analysis model between the influencing factors and the accident distance includes:

[0036] Constructing a convolution layer between the influencing factors and the accident distance;

[0037] The attention mechanism between the influencing factors and the accident distance is constructed using the following formula:

[0038]

[0039] in, represents the attention mechanism, d represents the dimension of the input data, Q represents the query matrix, K represents the key matrix, V represents the value matrix, Softmax represents the activation function, and Attention(Q,K,V) represents the output data of the attention mechanism;

[0040] The distillation mechanism between the influencing factors and the accident distance is constructed using the following formula:

[0041] y'=MaxPool(ELU(Conv(y)))

[0042] Where y'=MaxPool(ELU(Conv(y))) represents the distillation mechanism, y represents the input data of the distillation mechanism, Conv represents convolution, ELU represents the activation function used, MaxPool represents maximum pooling, and y' represents the output data of the distillation mechanism;

[0043] A convolutional neural network is used to determine the influencing factors and the accident distance using the convolutional layer;

[0044] Connecting the attention mechanism with the distillation mechanism to obtain a connection mechanism;

[0045] constructing a fully connected layer at the output of the connection mechanism;

[0046] The convolutional neural network, the connection mechanism and the fully connected layer are connected to obtain a distance analysis model.

[0047] Optionally, the using the safety distance and the historical safety log to generate the initial factors and safety values ​​between the historical high-voltage line and the historical operating personnel includes:

[0048] Arranging the safety distances to obtain a distance sequence;

[0049] Extracting security keywords and security values ​​of the security keywords from the historical security log;

[0050] Obtaining the internal factors of the high-voltage line, the internal factors of the personnel and the external factors of the environment between the historical high-voltage line and the historical operating personnel;

[0051] Searching for factor keywords belonging to the high-voltage line internal factors, the personnel internal factors and the environmental external factors from the safety keywords;

[0052] Acquire the factor value corresponding to the factor keyword from the security value;

[0053] Arranging the factor values ​​belonging to the same factor keyword among the factor values ​​to obtain a factor sequence;

[0054] performing factor preprocessing on the factor sequence to obtain preprocessing factors;

[0055] Generate initial factors between the historical high-voltage line and the historical operating personnel using the factor keywords, the preprocessing factors and the distance sequence;

[0056] Determining the safety index between the historical high-voltage line and the historical operating personnel and the index value of the safety index from the historical safety log;

[0057] The security index is weighted by using a preset entropy weight method to obtain the index weight of the security index;

[0058] The indicator weight is used to calculate the security value corresponding to the indicator value.

[0059] Optionally, analyzing the factor-safety correlation between the initial factor and the safety value includes:

[0060] The factor-safety correlation coefficient between the initial factor and the safety value is calculated using the following formula:

[0061]

[0062] Among them, ζ' i' (k') represents the factor-safety correlation coefficient, x' i' (k') represents the k'th value of the data of the ith category in the initial factor, x'0(k') represents the k'th value of the security value, x' j' (k') means dividing by x' i' (k') is the k'th value of the data of the j'th category outside (k'), α represents a constant, min k'The symbol for taking the minimum value, max k' The symbol for taking the maximum value;

[0063] According to the factor-safety association coefficient, the factor-safety association degree between the initial factor and the safety value is calculated using the following formula:

[0064]

[0065] Among them, r' 0i' represents the factor-safety correlation between the initial factor and the safety value, ζ' i' (k') represents the factor-safety association coefficient, and m' represents the number of k's.

[0066] Optionally, the constructing a safety warning model between the safety factor and the safety value includes:

[0067] Constructing an untrained safety warning model between the safety factor and the safety value;

[0068] Inputting the safety factor into the untrained safety warning model to obtain an output result;

[0069] Calculating a loss value between the security value and the output result;

[0070] According to the loss value, optimizing the parameters of the untrained safety warning model to obtain a safety warning model;

[0071] The untrained safety warning model includes an input layer, a hidden layer and an output layer, and the hidden layer includes:

[0072]

[0073] in, represents the hidden layer, X l (q) represents the qth value of the safety factor in the lth category, w l1 Represents X l (q) weight, L represents the number of categories of safety factors, b l1 Represents X l (q) bias, f represents the linear mapping function;

[0074] The output layer includes:

[0075]

[0076] in, represents the hidden layer, h p1 (q) represents the output of the hidden layer, w p2 Indicates hp1 (q), Q represents the number of connections between the hidden layer and the output layer, b p2 Indicates h p1 (q), f represents the linear mapping function, and p represents the connection number between the hidden layer and the output layer.

[0077] Optionally, optimizing parameters of the untrained safety warning model according to the loss value to obtain the safety warning model includes:

[0078] Calculating the partial derivative of the loss value with respect to the first neuron of the output layer in the untrained safety warning model;

[0079] Calculating the partial derivative of the loss value with respect to the second neuron of the hidden layer in the untrained safety warning model;

[0080] Using the first neuron partial derivative as a first error value between a weight parameter and a bias parameter of an output layer;

[0081] Correcting the weight parameter and the bias parameter of the output layer using the first error value to obtain a first correction parameter;

[0082] Using the second neuron partial derivative as a second error value between a weight parameter and a bias parameter of a hidden layer;

[0083] Correcting the weight parameter and the bias parameter of the hidden layer using the second error value to obtain a second correction parameter;

[0084] A safety warning model is determined by using the first correction parameter and the second correction parameter.

[0085] Optionally, outputting the security level corresponding to the security indicator through the security early warning model includes:

[0086] Obtaining the input layer, hidden layer and output layer in the security early warning model;

[0087] After receiving the security indicator through the input layer, the security indicator is transmitted to the hidden layer;

[0088] Using the hidden layer to extract hidden features from the security indicator to obtain hidden features;

[0089] According to the output layer, the implicit feature is mapped to each security category of the security warning model to obtain a mapping probability value under each security category;

[0090] A maximum probability value is selected from the mapped probability values, and a security category corresponding to the maximum probability value is used as the security level corresponding to the security indicator.

[0091] In order to solve the above problems, the present invention also provides a safety reminder system for high-altitude workers based on working voltage, the system comprising:

[0092] A factor selection module is used to collect historical accident logs and working voltages of historical high-voltage lines, screen the accident distance between the historical high-voltage line and historical operators and preliminary factors of the accident distance from the historical accident logs, splice the working voltage and the preliminary factors to obtain splicing factors, analyze the grey correlation between the splicing factors and the accident distance, and select the influencing factors of the accident distance from the splicing factors according to the grey correlation;

[0093] A distance output module is used to construct a distance analysis model between the influencing factors and the accident distance, collect the influencing index between the current high-voltage line and the current operator, input the influencing index into the distance analysis model, and output the safe distance corresponding to the influencing index through the distance analysis model;

[0094] A factor extraction module, used to collect the historical safety logs of the historical high-voltage line, generate the initial factors and safety values ​​between the historical high-voltage line and the historical operating personnel by using the safety distance and the historical safety logs, analyze the factor-safety correlation between the initial factors and the safety values, and extract the safety factors of the safety values ​​from the initial factors by using the factor-safety correlation;

[0095] A level output module is used to construct a safety warning model between the safety factor and the safety value, collect the safety index between the current high-voltage line and the current operator, input the safety index into the safety warning model, and output the safety level corresponding to the safety index through the safety warning model;

[0096] The safety prompt module is used to notify the current operator of the safety level to complete the safety prompt for the current operator and obtain the safety prompt result for the current operator.

[0097] Compared with the problems described in the background technology, the embodiment of the present invention selects the accident distance between the historical high-voltage line and the historical operating personnel and the preliminary factors of the accident distance from the historical accident log to select the factors that can affect the safety distance between the high-voltage line and the maintenance personnel. The embodiment of the present invention analyzes the gray correlation between the splicing factor and the accident distance to evaluate which splicing factors will affect the safety distance through the gray correlation. The embodiment of the present invention constructs a distance analysis model between the influencing factor and the accident distance to output the dynamic safety distance under the dynamic transformation environment using the distance analysis model. The embodiment of the present invention generates the initial factors and safety values ​​between the historical high-voltage line and the historical operating personnel by using the safety distance and the historical safety log to analyze the factors related to safety. The embodiment of the present invention constructs a safety warning model between the safety factor and the safety value to use the neural network model to warn the safety under different influencing factors. Therefore, the high-altitude worker safety prompt method based on working voltage proposed by the present invention can perform safety warning under the premise of dynamic safety distance. BRIEF DESCRIPTION OF THE DRAWINGS

[0098] Figure 1 A schematic flow chart of a method for providing safety reminders for high-altitude workers based on working voltage according to an embodiment of the present invention;

[0099] Figure 2 A schematic diagram of a module for implementing the method for providing safety reminders for high-altitude workers based on working voltage provided in one embodiment of the present invention.

[0100] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings in conjunction with the embodiments. DETAILED DESCRIPTION

[0101] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0102] The embodiment of the present application provides a method for providing safety reminders for high-altitude workers based on working voltage. The execution subject of the method for providing safety reminders for high-altitude workers based on working voltage includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided by the embodiment of the present application. In other words, the method for providing safety reminders for high-altitude workers based on working voltage can be executed by software or hardware installed in a terminal device or a server device. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0103] Embodiment 1:

[0104] Reference Figure 1FIG. 1 is a flow chart of a method for providing a safety reminder for high-altitude workers based on working voltage according to an embodiment of the present invention. In this embodiment, the method for providing a safety reminder for high-altitude workers based on working voltage includes:

[0105] S1. Collect historical accident logs and working voltages of historical high-voltage lines, select accident distances between the historical high-voltage lines and historical operators and preliminary factors of the accident distance from the historical accident logs, splice the working voltage and the preliminary factors to obtain splicing factors, analyze the grey correlation between the splicing factors and the accident distance, and select influencing factors of the accident distance from the splicing factors according to the grey correlation.

[0106] In the embodiment of the present invention, the historical high-voltage line refers to a high-voltage line where an accident occurred in a historical period, the historical accident log refers to a record of accidents occurring on the high-voltage line, and the working voltage refers to the voltage when the historical high-voltage line accident occurred.

[0107] Furthermore, the embodiment of the present invention selects the accident distance between the historical high-voltage line and the historical operating personnel and the preliminary factors of the accident distance from the historical accident log to select the factors that can affect the safe distance between the high-voltage line and the maintenance personnel.

[0108] The preliminary factors are factors that can affect the safe distance between the high-voltage line and maintenance personnel.

[0109] In one embodiment of the present invention, the screening of the accident distance between the historical high-voltage line and the historical operating personnel and the preliminary factors of the accident distance from the historical accident log includes: querying the accident address between the historical high-voltage line and the historical operating personnel from the historical accident log; clustering the accident address to obtain the cluster address; identifying the address center of the cluster address; calculating the initial distance between the position of the historical high-voltage line and the address center; performing distance preprocessing on the initial distance to obtain the accident distance; extracting accident keywords and accident values ​​of the accident keywords from the historical accident log; dividing the high-voltage line internal factors, personnel internal factors and environmental external factors between the historical high-voltage line and the historical operating personnel; querying the target keywords belonging to the high-voltage line internal factors, the personnel internal factors and the environmental external factors from the accident keywords; obtaining the target value corresponding to the target keyword from the accident value; arranging the target values ​​belonging to the same target keyword in the target values ​​to obtain a target sequence; performing target preprocessing on the target sequence to obtain a preprocessing sequence; and generating the preliminary factors of the accident distance using the target keyword and the preprocessing sequence.

[0110] Among them, the clustered address refers to a collection of similar addresses, the accident keyword refers to the noun after removing useless subject, predicate, object and modal particles, the high-voltage line internal cause refers to all factors related to the high-voltage line, the personnel internal cause refers to all factors related to personnel, and the environmental external cause refers to all factors related to the environment.

[0111] Optionally, the process of performing distance preprocessing on the initial distance to obtain the accident distance refers to performing dimensionless processing on the initial distance, and further, the process of performing target preprocessing on the target sequence to obtain the preprocessed sequence refers to performing dimensionless processing on the target sequence.

[0112] Optionally, the process of splicing the operating voltage and the preliminary factors refers to including the operating voltage and the preliminary factors as data in a set. Furthermore, the process of generating the preliminary factors of the accident distance using the target keywords and the preprocessing sequence refers to using the preprocessing sequence corresponding to each target keyword as a preliminary factor.

[0113] Furthermore, the embodiment of the present invention analyzes the grey correlation between the splicing factors and the accident distance to evaluate which splicing factors will affect the safety distance through the grey correlation.

[0114] In one embodiment of the present invention, the analyzing the grey correlation between the splicing factor and the accident distance includes: obtaining the working voltage and the preliminary factor in the splicing factor; performing voltage preprocessing on the working voltage to obtain a preprocessing voltage; splicing the preprocessing voltage and the preliminary factor to obtain a splicing result; and calculating the grey correlation coefficient between the splicing result and the accident distance using the following formula:

[0115]

[0116] Among them, i (k) represents the grey relational coefficient, x i (k) represents the kth value of the data of the ith category in the splicing result, x0(k) represents the kth cluster value of the accident distance, x j (k) means division by x i (k) is the kth value of the data of the jth category outside the range, ρ represents a constant, min k The symbol for taking the minimum value, max k The symbol for taking the maximum value;

[0117] According to the grey correlation coefficient, the grey correlation degree between the splicing result and the accident distance is calculated using the following formula:

[0118]

[0119] Among them, r 0i represents the grey correlation between the ith target keyword and the accident distance, ζ i (k) represents the grey relational coefficient, and m represents the number of k.

[0120] Optionally, the process of performing voltage preprocessing on the working voltage refers to the process of performing dimensionless processing on the working voltage.

[0121] Optionally, the process of selecting the influencing factors of the accident distance from the splicing factors according to the grey correlation degree refers to: setting a threshold of the grey correlation degree, and taking the factors corresponding to the grey correlation degree that meets the threshold as influencing factors.

[0122] S2. Construct a distance analysis model between the influencing factors and the accident distance, collect the influencing indicators between the current high-voltage line and the current operating personnel, input the influencing indicators into the distance analysis model, and output the safe distance corresponding to the influencing indicators through the distance analysis model.

[0123] The embodiment of the present invention constructs a distance analysis model between the influencing factors and the accident distance, so as to utilize the distance analysis model to output a dynamic safety distance in a dynamic transformation environment.

[0124] In one embodiment of the present invention, the construction of the distance analysis model between the influencing factors and the accident distance includes: constructing a convolution layer between the influencing factors and the accident distance; and constructing an attention mechanism between the influencing factors and the accident distance using the following formula:

[0125]

[0126] in, represents the attention mechanism, d represents the dimension of the input data, Q represents the query matrix, K represents the key matrix, V represents the value matrix, Softmax represents the activation function, and Attention(Q,K,V) represents the output data of the attention mechanism;

[0127] The distillation mechanism between the influencing factors and the accident distance is constructed using the following formula:

[0128] y'=MaxPool(ELU(Conv(y)))

[0129] Where y'=MaxPool(ELU(Conv(y))) represents the distillation mechanism, y represents the input data of the distillation mechanism, Conv represents convolution, ELU represents the activation function used, MaxPool represents maximum pooling, and y' represents the output data of the distillation mechanism;

[0130] The convolutional layer is used to determine the convolutional neural network between the influencing factors and the accident distance; the attention mechanism is connected to the distillation mechanism to obtain a connection mechanism; a fully connected layer is constructed at the output end of the connection mechanism; the convolutional neural network, the connection mechanism and the fully connected layer are connected to obtain a distance analysis model.

[0131] It should be noted that the output end of the attention mechanism is the distillation mechanism, the output end of the distillation mechanism is the fully connected layer, and the output end of the convolutional neural network is the attention mechanism.

[0132] Furthermore, in an embodiment of the present invention, the influencing index refers to input data of a distance analysis model between a current high-voltage line and a current operator.

[0133] S3. Collect the historical safety logs of the historical high-voltage line, use the safety distance and the historical safety logs to generate initial factors and safety values ​​between the historical high-voltage line and the historical operating personnel, analyze the factor-safety correlation between the initial factors and the safety values, and use the factor-safety correlation to extract the safety factors of the safety values ​​from the initial factors.

[0134] In the embodiment of the present invention, the historical safety log refers to a record of unsafe events that occurred to operators during a historical period.

[0135] The embodiment of the present invention generates initial factors and safety values ​​between the historical high-voltage line and the historical operators by using the safety distance and the historical safety log, so as to analyze the factors related to safety.

[0136] The initial factors refer to factors related to security, and the security value refers to a value for evaluating the security level.

[0137] In one embodiment of the present invention, the use of the safety distance and the historical safety log to generate the initial factors and safety values ​​between the historical high-voltage line and the historical operating personnel includes: arranging the safety distance to obtain a distance sequence; extracting safety keywords and the safety values ​​of the safety keywords from the historical safety log; obtaining the high-voltage line internal factors, personnel internal factors and environmental external factors between the historical high-voltage line and the historical operating personnel; querying the factor keywords belonging to the high-voltage line internal factors, the personnel internal factors and the environmental external factors from the safety keywords; obtaining the factor values ​​corresponding to the factor keywords from the safety values; arranging the factor values ​​belonging to the same factor keyword in the factor values ​​to obtain a factor sequence; preprocessing the factor sequence to obtain preprocessed factors; generating the initial factors between the historical high-voltage line and the historical operating personnel using the factor keywords, the preprocessing factors and the distance sequence; determining the safety index between the historical high-voltage line and the historical operating personnel and the index value of the safety index from the historical safety log; using a preset entropy weight method to assign index weights to the safety index to obtain the index weight of the safety index; and calculating the safety value corresponding to the index value using the index weight.

[0138] The safety index refers to whether casualties occur, including the number of injured persons, the number of deaths, etc.

[0139] Optionally, the process of generating the initial factors between the historical high-voltage lines and historical operating personnel using the factor keywords, the preprocessing factors and the distance sequence refers to taking the preprocessing factors and the distance sequence corresponding to each factor keyword as the initial factors.

[0140] In one embodiment of the present invention, the analyzing the factor-safety association degree between the initial factor and the safety value includes: calculating the factor-safety association coefficient between the initial factor and the safety value using the following formula:

[0141]

[0142] Among them, ζ' i' (k') represents the factor-safety correlation coefficient, x' i' (k') represents the k'th value of the data of the ith category in the initial factor, x'0(k') represents the k'th value of the security value, x' j' (k') means dividing by x' i' (k') is the k'th value of the data of the j'th category outside (k'), α represents a constant, min k' The symbol for taking the minimum value, max k' The symbol for taking the maximum value;

[0143] According to the factor-safety association coefficient, the factor-safety association degree between the initial factor and the safety value is calculated using the following formula:

[0144]

[0145] Among them, r' 0i' represents the factor-safety correlation between the initial factor and the safety value, ζ' i' (k') represents the factor-safety association coefficient, and m' represents the number of k's.

[0146] S4. Construct a safety warning model between the safety factors and the safety values, collect safety indicators between the current high-voltage line and the current operating personnel, input the safety indicators into the safety warning model, and output the safety level corresponding to the safety indicators through the safety warning model.

[0147] The embodiment of the present invention constructs a safety warning model between the safety factor and the safety value, so as to use a neural network model to warn of safety under different influencing factors.

[0148] In one embodiment of the present invention, the construction of the security warning model between the security factor and the security value includes: constructing an untrained security warning model between the security factor and the security value; inputting the security factor into the untrained security warning model to obtain an output result; calculating a loss value between the security value and the output result; optimizing parameters of the untrained security warning model according to the loss value to obtain a security warning model; wherein the untrained security warning model includes an input layer, a hidden layer and an output layer, and the hidden layer includes:

[0149]

[0150] in, represents the hidden layer, X l (q) represents the qth value of the safety factor in the lth category, w l1 Represents X l (q) weight, L represents the number of categories of safety factors, b l1 Represents X l (q) bias, f represents the linear mapping function;

[0151] The output layer includes:

[0152]

[0153] in, represents the hidden layer, h p1 (q) represents the output of the hidden layer, w p2 Indicates h p1 (q), Q represents the number of connections between the hidden layer and the output layer, b p2 Indicates h p1 (q), f represents the linear mapping function, and p represents the connection number between the hidden layer and the output layer.

[0154] In another embodiment of the present invention, the parameter optimization of the untrained safety warning model is performed according to the loss value to obtain the safety warning model, including: calculating the first neuron partial derivative of the loss value with respect to the output layer in the untrained safety warning model; calculating the second neuron partial derivative of the loss value with respect to the hidden layer in the untrained safety warning model; using the first neuron partial derivative as the first error value of the weight parameter and bias parameter of the output layer; using the first error value to correct the weight parameter and bias parameter of the output layer to obtain the first correction parameter; using the second neuron partial derivative as the second error value of the weight parameter and bias parameter of the hidden layer; using the second error value to correct the weight parameter and bias parameter of the hidden layer to obtain the second correction parameter; and determining the safety warning model through the first correction parameter and the second correction parameter.

[0155] Optionally, the process of using the first error value to correct the weight parameters and bias parameters of the output layer to obtain the first correction parameters refers to adding the weight parameters and bias parameters of the output layer to the first error value respectively, thereby obtaining the corrected values ​​of the weight parameters and bias parameters of the output layer.

[0156] Furthermore, in the embodiment of the present invention, the security level refers to the level for evaluating whether the security is verified, including normal and abnormal, and the abnormal level includes mild, moderate and severe, etc.

[0157] In one embodiment of the present invention, the security level corresponding to the security indicator outputted through the security warning model includes: obtaining the input layer, hidden layer and output layer in the security warning model; after receiving the security indicator through the input layer, passing the security indicator to the hidden layer; using the hidden layer to extract implicit features of the security indicator to obtain implicit features; according to the output layer, mapping the implicit features to each security category of the security warning model to obtain a mapping probability value under each security category; selecting the maximum probability value from the mapped probability values ​​and using the security category corresponding to the maximum probability value as the security level corresponding to the security indicator.

[0158] S5. Notify the current operator of the safety level to complete the safety prompt for the current operator and obtain the safety prompt result for the current operator.

[0159] Optionally, the process of notifying the current operator of the safety level may be implemented through a voice signal or a text display on a screen.

[0160] Compared with the problems described in the background technology, the embodiment of the present invention selects the accident distance between the historical high-voltage line and the historical operating personnel and the preliminary factors of the accident distance from the historical accident log to select the factors that can affect the safety distance between the high-voltage line and the maintenance personnel. The embodiment of the present invention analyzes the gray correlation between the splicing factor and the accident distance to evaluate which splicing factors will affect the safety distance through the gray correlation. The embodiment of the present invention constructs a distance analysis model between the influencing factor and the accident distance to output the dynamic safety distance under the dynamic transformation environment using the distance analysis model. The embodiment of the present invention generates the initial factors and safety values ​​between the historical high-voltage line and the historical operating personnel by using the safety distance and the historical safety log to analyze the factors related to safety. The embodiment of the present invention constructs a safety warning model between the safety factor and the safety value to use the neural network model to warn the safety under different influencing factors. Therefore, the high-altitude worker safety prompt method based on working voltage proposed by the present invention can perform safety warning under the premise of dynamic safety distance.

[0161] Embodiment 2:

[0162] like Figure 2 The figure shows a functional module diagram of a safety reminder system for high-altitude workers based on working voltage according to the present invention.

[0163] The high-altitude worker safety reminder system 200 based on working voltage of the present invention can be installed in an electronic device. According to the functions to be implemented, the high-altitude worker safety reminder system based on working voltage can include a factor selection module 201, a distance output module 202, a factor extraction module 203, a level output module 204 and a safety reminder module 205. The module of the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0164] In the embodiment of the present invention, the functions of each module / unit are as follows:

[0165] The factor selection module 201 is used to collect historical accident logs and working voltages of historical high-voltage lines, screen the accident distance between the historical high-voltage line and historical operators and the preliminary factors of the accident distance from the historical accident logs, splice the working voltage and the preliminary factors to obtain splicing factors, analyze the grey correlation between the splicing factors and the accident distance, and select the influencing factors of the accident distance from the splicing factors according to the grey correlation;

[0166] The distance output module 202 is used to construct a distance analysis model between the influencing factors and the accident distance, collect the influencing index between the current high-voltage line and the current operator, input the influencing index into the distance analysis model, and output the safe distance corresponding to the influencing index through the distance analysis model;

[0167] The factor extraction module 203 is used to collect the historical safety logs of the historical high-voltage line, generate the initial factors and safety values ​​between the historical high-voltage line and the historical operators by using the safety distance and the historical safety logs, analyze the factor-safety correlation between the initial factors and the safety values, and extract the safety factors of the safety values ​​from the initial factors by using the factor-safety correlation;

[0168] The level output module 204 is used to construct a safety warning model between the safety factor and the safety value, collect the safety index between the current high-voltage line and the current operator, input the safety index into the safety warning model, and output the safety level corresponding to the safety index through the safety warning model;

[0169] The safety prompt module 205 is used to notify the current operator of the safety level to complete the safety prompt for the current operator and obtain the safety prompt result for the current operator.

[0170] In detail, each module in the high-altitude worker safety reminder system 200 based on working voltage in the embodiment of the present invention is used in the same manner as described above. Figure 1 The technical means are the same as the safety reminder method for high-altitude workers based on working voltage described in, and can produce the same technical effects, so they will not be repeated here.

[0171] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0172] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A safety reminder method for high-altitude workers based on working voltage, characterized in that: The method comprises: Collect historical accident logs and working voltages of historical high-voltage lines, select accident distances between the historical high-voltage lines and historical operators and preliminary factors of the accident distance from the historical accident logs, splice the working voltage and the preliminary factors to obtain splicing factors, analyze the grey correlation between the splicing factors and the accident distance, and select influencing factors of the accident distance from the splicing factors according to the grey correlation; Constructing a distance analysis model between the influencing factors and the accident distance, collecting the influencing index between the current high-voltage line and the current operator, inputting the influencing index into the distance analysis model, and outputting the safe distance corresponding to the influencing index through the distance analysis model; Collecting historical safety logs of the historical high-voltage line, generating initial factors and safety values ​​between the historical high-voltage line and historical operators using the safety distance and the historical safety logs, analyzing the factor-safety correlation between the initial factors and the safety values, and extracting the safety factors of the safety values ​​from the initial factors using the factor-safety correlation; Constructing a safety warning model between the safety factor and the safety value, collecting the safety index between the current high-voltage line and the current operator, inputting the safety index into the safety warning model, and outputting the safety level corresponding to the safety index through the safety warning model; The safety level is notified to the current operator to complete the safety prompt for the current operator and obtain the safety prompt result for the current operator.

2. The method for providing safety reminders for high-altitude workers based on working voltage according to claim 1, characterized in that: The screening of the accident distance between the historical high-voltage line and the historical operator and the preliminary factors of the accident distance from the historical accident log includes: Querying the accident address between the historical high-voltage line and the historical operator from the historical accident log; Performing address clustering on the accident address to obtain a clustered address; identifying an address center of the clustered addresses; Calculating an initial distance between the location of the historical high-voltage line and the center of the address; Performing distance preprocessing on the initial distance to obtain the accident distance; Extracting accident keywords and accident values ​​of the accident keywords from the historical accident log; Divide the high-voltage line internal factors, personnel internal factors and environmental external factors between the historical high-voltage line and the historical operating personnel; Searching for target keywords belonging to the internal cause of the high-voltage line, the internal cause of the personnel, and the external cause of the environment from the accident keywords; Acquire a target value corresponding to the target keyword from the accident value; Arranging the target values ​​belonging to the same target keyword among the target values ​​to obtain a target sequence; Performing target preprocessing on the target sequence to obtain a preprocessed sequence; The target keyword and the preprocessing sequence are used to generate preliminary factors of the accident distance.

3. The method for providing safety reminders for high-altitude workers based on working voltage according to claim 1, characterized in that: The analyzing the grey correlation between the splicing factor and the accident distance includes: Obtaining the working voltage and preliminary factors in the splicing factors; Performing voltage preprocessing on the working voltage to obtain a preprocessing voltage; splicing the preprocessing voltage and the preliminary factor to obtain a splicing result; The grey correlation coefficient between the splicing result and the accident distance is calculated using the following formula: Among them, i (k) represents the grey relational coefficient, x i (k) represents the kth value of the data of the ith category in the splicing result, x0(k) represents the kth cluster value of the accident distance, x j (k) means division by x i (k) is the kth value of the data of the jth category outside the range, ρ represents a constant, min k The symbol for taking the minimum value, max k The symbol for taking the maximum value; According to the grey correlation coefficient, the grey correlation degree between the splicing result and the accident distance is calculated using the following formula: Among them, r 0i represents the grey correlation between the ith target keyword and the accident distance, ζ i (k) represents the grey relational coefficient, and m represents the number of k.

4. The method for providing safety reminders for high-altitude workers based on working voltage according to claim 1, characterized in that: The constructing of the distance analysis model between the influencing factors and the accident distance includes: Constructing a convolution layer between the influencing factors and the accident distance; The attention mechanism between the influencing factors and the accident distance is constructed using the following formula: in, represents the attention mechanism, d represents the dimension of the input data, Q represents the query matrix, K represents the key matrix, V represents the value matrix, Softmax represents the activation function, and Attention(Q,K,V) represents the output data of the attention mechanism; The distillation mechanism between the influencing factors and the accident distance is constructed using the following formula: y'=MaxPool(ELU(Conv(y))) Where y'=MaxPool(ELU(Conv(y))) represents the distillation mechanism, y represents the input data of the distillation mechanism, Conv represents convolution, ELU represents the activation function used, MaxPool represents maximum pooling, and y' represents the output data of the distillation mechanism; A convolutional neural network is used to determine the influencing factors and the accident distance using the convolutional layer; Connecting the attention mechanism with the distillation mechanism to obtain a connection mechanism; constructing a fully connected layer at the output of the connection mechanism; The convolutional neural network, the connection mechanism and the fully connected layer are connected to obtain a distance analysis model.

5. The method for providing safety reminders for high-altitude workers based on working voltage according to claim 1, characterized in that: The method of using the safety distance and the historical safety log to generate the initial factors and safety values ​​between the historical high-voltage line and the historical operating personnel includes: Arranging the safety distances to obtain a distance sequence; Extracting security keywords and security values ​​of the security keywords from the historical security log; Obtaining the internal factors of the high-voltage line, the internal factors of the personnel and the external factors of the environment between the historical high-voltage line and the historical operating personnel; Searching for factor keywords belonging to the high-voltage line internal factors, the personnel internal factors and the environmental external factors from the safety keywords; Acquire the factor value corresponding to the factor keyword from the security value; Arranging the factor values ​​belonging to the same factor keyword among the factor values ​​to obtain a factor sequence; performing factor preprocessing on the factor sequence to obtain preprocessing factors; Generate initial factors between the historical high-voltage line and the historical operating personnel using the factor keywords, the preprocessing factors and the distance sequence; Determining the safety index between the historical high-voltage line and the historical operating personnel and the index value of the safety index from the historical safety log; The security index is weighted by using a preset entropy weight method to obtain the index weight of the security index; The indicator weight is used to calculate the security value corresponding to the indicator value.

6. The method for providing safety reminders for high-altitude workers based on working voltage according to claim 1, characterized in that: The analyzing the factor-safety correlation between the initial factor and the safety value includes: The factor-safety correlation coefficient between the initial factor and the safety value is calculated using the following formula: Among them, ζ' i' (k') represents the factor-safety correlation coefficient, x' i' (k') represents the k'th value of the data of the ith category in the initial factor, x'0(k') represents the k'th value of the security value, x' j' (k') means dividing by x' i' (k') is the k'th value of the data of the j'th category outside (k'), α represents a constant, min k' The symbol for taking the minimum value, max k' The symbol for taking the maximum value; According to the factor-safety association coefficient, the factor-safety association degree between the initial factor and the safety value is calculated using the following formula: Among them, r' 0i' represents the factor-safety correlation between the initial factor and the safety value, ζ' i' (k') represents the factor-safety association coefficient, and m' represents the number of k's.

7. The method for providing safety reminders for high-altitude workers based on working voltage according to claim 1, characterized in that: The constructing of a safety early warning model between the safety factor and the safety value includes: Constructing an untrained safety warning model between the safety factor and the safety value; Inputting the safety factor into the untrained safety warning model to obtain an output result; Calculating a loss value between the security value and the output result; According to the loss value, optimizing the parameters of the untrained safety warning model to obtain a safety warning model; The untrained safety warning model includes an input layer, a hidden layer and an output layer, and the hidden layer includes: in, represents the hidden layer, X l (q) represents the qth value of the safety factor in the lth category, w l1 Represents X l (q) weight, L represents the number of categories of safety factors, b l1 Represents X l (q) bias, f represents the linear mapping function; The output layer includes: in, represents the hidden layer, h p1 (q) represents the output of the hidden layer, w p2 Indicates h p1 (q), Q represents the number of connections between the hidden layer and the output layer, b p2 Indicates h p1 (q), f represents the linear mapping function, and p represents the connection number between the hidden layer and the output layer.

8. The method for providing safety reminders for high-altitude workers based on working voltage according to claim 1, characterized in that: The step of optimizing parameters of the untrained safety warning model according to the loss value to obtain a safety warning model includes: Calculating the partial derivative of the loss value with respect to the first neuron of the output layer in the untrained safety warning model; Calculating the partial derivative of the loss value with respect to the second neuron of the hidden layer in the untrained safety warning model; Using the first neuron partial derivative as a first error value between a weight parameter and a bias parameter of an output layer; Correcting the weight parameter and the bias parameter of the output layer using the first error value to obtain a first correction parameter; Using the second neuron partial derivative as a second error value between a weight parameter and a bias parameter of a hidden layer; Correcting the weight parameter and the bias parameter of the hidden layer using the second error value to obtain a second correction parameter; A safety warning model is determined by using the first correction parameter and the second correction parameter.

9. The method for providing safety reminders for high-altitude workers based on working voltage according to claim 1, characterized in that: Outputting the security level corresponding to the security indicator through the security early warning model includes: Obtaining the input layer, hidden layer and output layer in the security early warning model; After receiving the security indicator through the input layer, the security indicator is transmitted to the hidden layer; Using the hidden layer to extract hidden features of the security indicator to obtain hidden features; According to the output layer, the implicit feature is mapped to each security category of the security warning model to obtain a mapping probability value under each security category; A maximum probability value is selected from the mapped probability values, and a security category corresponding to the maximum probability value is used as the security level corresponding to the security indicator.

10. A safety reminder system for high-altitude workers based on working voltage, characterized in that: The system comprises: A factor selection module is used to collect historical accident logs and working voltages of historical high-voltage lines, screen the accident distance between the historical high-voltage line and historical operators and preliminary factors of the accident distance from the historical accident logs, splice the working voltage and the preliminary factors to obtain splicing factors, analyze the grey correlation between the splicing factors and the accident distance, and select the influencing factors of the accident distance from the splicing factors according to the grey correlation; A distance output module is used to construct a distance analysis model between the influencing factors and the accident distance, collect the influencing index between the current high-voltage line and the current operator, input the influencing index into the distance analysis model, and output the safe distance corresponding to the influencing index through the distance analysis model; A factor extraction module, used to collect the historical safety logs of the historical high-voltage line, generate the initial factors and safety values ​​between the historical high-voltage line and the historical operating personnel by using the safety distance and the historical safety logs, analyze the factor-safety correlation between the initial factors and the safety values, and extract the safety factors of the safety values ​​from the initial factors by using the factor-safety correlation; A level output module is used to construct a safety warning model between the safety factor and the safety value, collect the safety index between the current high-voltage line and the current operator, input the safety index into the safety warning model, and output the safety level corresponding to the safety index through the safety warning model; The safety prompt module is used to notify the current operator of the safety level to complete the safety prompt for the current operator and obtain the safety prompt result for the current operator.

Images