AI model algorithm for power transmission line affected by natural conditions

Abstract

The invention relates to the technical field of power systems, in particular to an AI model algorithm for a power transmission line affected by natural conditions. The method comprises the following steps: a data acquisition and processing unit acquires wind speed, temperature, humidity and circuit load data, extracts time sequence features by adopting a time sequence convolutional neural network fusion model, and fuses the time sequence features by using an attention mechanism to obtain fusion features; the environmental factor modeling unit introduces a time-varying dynamic correction factor, and establishes a power transmission line temperature and humidity-stress influence model by using a nonlinear regression model; the real-time analysis and correction unit adjusts working parameters of the power transmission line based on a power transmission line temperature and humidity-stress influence model in combination with circuit load data; and the data visualization unit displays the environmental data and the power transmission line temperature and humidity-stress influence model through a digital twin management platform. According to the AI model algorithm for the power transmission line influenced by the natural conditions, time sequence feature modeling is carried out on environmental factors through a time sequence convolutional neural network and an attention mechanism, and the influence of the environment on the power transmission line is captured.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to an AI model algorithm for a transmission line affected by natural conditions (AI model algorithm for a transmission line affected by natural conditions). Background Art

[0002] An AI model algorithm for transmission lines affected by natural conditions aims to improve the operating safety of transmission lines under severe weather conditions and optimize power transmission efficiency. By introducing a time series convolutional neural network fusion model and a time-varying dynamic correction factor, the temperature, humidity and stress effects of the transmission lines and the circuit load distribution are controlled, and line parameters are adjusted dynamically in real time to avoid stress exceeding the limit caused by temperature and humidity changes and wind speed fluctuations, thereby achieving safe, stable and efficient operation of the transmission lines.

[0003] The existing AI model algorithm for transmission lines affected by natural conditions is usually difficult to accurately predict and regulate the dynamic changes of line stress caused by complex climatic conditions. In addition, since transmission lines are often exposed to the environment, they are greatly affected by the environmental factors such as wind, temperature and humidity, resulting in simple transmission line stress imbalance and uneven load distribution leading to aging and depreciation of transmission lines. Therefore, an AI model algorithm for transmission lines affected by natural conditions is designed. Summary of the invention

[0004] The purpose of the present invention is to provide an AI model algorithm for transmission lines affected by natural conditions, so as to solve the problem raised in the above background technology that transmission lines are often exposed to the environment and are greatly affected by the environmental influences of wind, temperature and humidity, resulting in simple transmission line stress imbalance and uneven load distribution leading to aging and depreciation of transmission lines, and at the same time learn and train the law of change of three-dimensional coordinates of transmission lines affected by natural conditions.

[0005] To achieve the above object, the present invention aims to provide an AI model algorithm for power transmission lines affected by natural conditions, comprising:

[0006] A data acquisition processing unit, wherein the data acquisition processing unit uses sensors to collect wind speed, temperature, humidity and circuit load data in real time, uses a time series convolutional neural network fusion model to perform time series modeling on wind speed, temperature and humidity to extract their time series features, uses an attention mechanism to fuse the wind speed, temperature and humidity time series features to obtain fusion features, and stores the fusion features in a time series database;

[0007] It also includes an environmental factor modeling unit, which uses a nonlinear regression model to establish a transmission line temperature, humidity and stress impact model based on the fusion characteristics of wind speed, temperature and humidity and introduces a time-varying dynamic correction factor;

[0008] It also includes a real-time analysis and correction unit, which dynamically adjusts the transmission line operating parameters based on the transmission line temperature, humidity and stress influence model combined with circuit load data;

[0009] It also includes a data processing unit, which displays wind speed, temperature, humidity, circuit load data and transmission line temperature and humidity-stress impact model data through a digital twin management platform.

[0010] As a further improvement of the technical solution, the data acquisition and processing unit includes a data acquisition module, a data processing module and a data storage module;

[0011] The data acquisition module uses a sensor to collect wind speed V in real time. wind (t), temperature T(t), humidity H(t) and circuit load data L(t);

[0012] Where, t is time;

[0013] The data processing module uses a time series convolutional neural network fusion model to perform time series modeling on wind speed, temperature and humidity to extract their time series features, and uses an attention mechanism to fuse the time series features of wind speed, temperature and humidity to obtain fusion features;

[0014] The data storage module is used to store the fusion features in a time series database;

[0015] The time series convolutional neural network fusion model is implemented based on the fusion of convolutional neural network and one-dimensional convolution, and the corresponding time series features are extracted by performing convolution operations on wind speed, temperature and humidity;

[0016] The attention mechanism is used to fuse the temporal features of wind speed, temperature and humidity.

[0017] As a further improvement of the technical solution, the data processing module uses a time series convolutional neural network fusion model to perform time series modeling on wind speed, temperature and humidity to extract their time series features, and uses an attention mechanism to fuse the time series features of wind speed, temperature and humidity to obtain fusion features. The specific method is as follows:

[0018] S1.2.1. Use the time series convolutional neural network fusion model to model the wind speed, temperature and humidity and extract their time series characteristics:

[0019] Wind speed time series characteristics:

[0020] X wind (t) = σ(W wind ·V wind (t)+b wind );

[0021] Temperature timing characteristics:

[0022] X temp (t) = σ(W temp ·T(t)+b temp );

[0023] Humidity time series characteristics:

[0024] X humidity (t) = σ(W humidity ·H(t)+b humidity );

[0025] Among them, X wind (t) is the time series characteristic of wind speed; X temp (t) is the temperature time series characteristic; X humidity (t) is the humidity time series feature; σ is the ReLU function; W wind is the wind speed convolution kernel; b wind is the wind speed bias term; W temp is the temperature convolution kernel; b temp is the temperature bias term; W humidity is the humidity convolution kernel; b humidity is the humidity bias term; t is the time;

[0026] S1.2.2. Use the attention mechanism to calculate wind speed score, temperature score and humidity score:

[0027] Wind Speed ​​Rating:

[0028] score wind (t) = f(W′ wind ·X wind (t)+b wind );

[0029] Temperature Rating:

[0030] score temp (t) = f(W′ temp ·X temp (t)+b temp );

[0031] Moisture Rating:

[0032] score humidity (t) = f(W′ humidity ·X humidity (t)+b humidity );

[0033] Among them, score wind (t) is the wind speed score; score temp (t) is the temperature score; score humidity(t) is the humidity score; W' wind is the wind speed weight matrix; W' temp is the temperature weight matrix; W' humidity is the humidity weight matrix; f is the Sigmoid activation function;

[0034] S1.2.3. Calculate the attention weights of wind speed, temperature and humidity based on wind speed score, temperature score and humidity score:

[0035] Wind speed attention weight:

[0036]

[0037] Temperature attention weight:

[0038]

[0039] Humidity attention weight:

[0040]

[0041] Among them, α wind (t) is the wind speed attention weight; α temp (t) is the temperature attention weight; α humidity (t) is the humidity attention weight;

[0042] S1.2.4. Based on the attention weights of wind speed, temperature and humidity, the time series features of wind speed, temperature and humidity are weighted and fused to obtain the fused features:

[0043] X fused (t) = α wind (t)·X wind (t)+α temp (t)·X temp (t)+α humidity (t)·X humidity (t);

[0044] Among them, X fused (t) is the fusion feature.

[0045] As a further improvement of the technical solution, the environmental factor modeling unit is based on the fusion characteristics of wind speed, temperature and humidity and introduces a time-varying dynamic correction factor, and uses a nonlinear regression model to establish a transmission line temperature and humidity-stress influence model. The specific method steps are as follows:

[0046] S2.1. Calculate the thermal expansion of the transmission line based on temperature, calculate the wind pressure of the transmission line based on wind speed, and calculate the humidity load of the transmission line based on humidity;

[0047] S2.2, define the time-varying dynamic correction factor;

[0048] S2.3. Based on the thermal expansion of the transmission line, the wind pressure of the transmission line, the humidity load of the transmission line and the time-varying dynamic correction factor, a nonlinear regression model is used to construct the temperature, humidity and stress influence model of the transmission line.

[0049] As a further improvement of the technical solution, in S2.1, the thermal expansion of the transmission line is calculated based on the temperature, the wind pressure of the transmission line is calculated based on the wind speed, and the humidity load of the transmission line is calculated based on the humidity. The specific method is as follows:

[0050] S2.1.1. Calculation of thermal expansion of transmission lines based on temperature:

[0051] ΔL=α1L0(T(t)-T0);

[0052] Among them, ΔL is the thermal expansion of the transmission line; α1 is the linear expansion coefficient of the transmission line material; L0 is the initial length of the transmission line; T0 is the initial temperature;

[0053] S2.1.2. Calculate the wind pressure on the transmission line based on wind speed:

[0054] P wind =C d ·ρ air ·A wire ·V wind (t) 2 ;

[0055] Among them, P wind is the wind pressure on the transmission line; C d is the drag coefficient; ρ air is the air density; A wire is the cross-sectional area of ​​the transmission line;

[0056] S2.1.3. Calculation of humidity load on transmission lines based on humidity

[0057] P humidity =k humidity ·A wire H(t);

[0058] Among them, P humidity is the humidity load of the transmission line; k humidity is the constant for humidity attachment.

[0059] As a further improvement of the technical solution, in S2.2, a time-varying dynamic correction factor is defined, and the specific method is as follows:

[0060]

[0061] Among them, β(t) is the time-varying dynamic correction factor; γ1 and γ2 are constants of the correction factor; λ1 is the attenuation rate factor; λ2 is the periodic factor; δ is the phase factor.

[0062] As a further improvement of the technical solution, in S2.3, based on the thermal expansion of the transmission line, the wind pressure of the transmission line, the humidity load of the transmission line and the time-varying dynamic correction factor, a nonlinear regression model is used to construct a temperature, humidity and stress influence model of the transmission line. The specific method is as follows:

[0063] Transmission line temperature, humidity and stress impact model:

[0064] S(t)=θ0+θ1·ΔL+θ2·P wind +θ3·P humidity +β(t)·X fused (t);

[0065] Among them, θ0 is the constant coefficient; θ1 is the weight of thermal expansion of the transmission line; θ2 is the weight of wind pressure of the transmission line; θ3 is the weight of humidity load of the transmission line; S(t) is the stress of the transmission line.

[0066] As a further improvement of the technical solution, the real-time analysis and correction unit includes a model analysis module and a parameter correction module;

[0067] The model analysis module calculates the stress of the transmission line based on the temperature, humidity and stress influence model of the transmission line, and determines whether it is within a safe range;

[0068] The parameter correction module is used to correct the transmission line operating parameters in real time when the transmission line stress is not within a safe range.

[0069] As a further improvement of the technical solution, the model analysis module calculates the stress of the transmission line based on the temperature and humidity-stress influence model of the transmission line, and determines whether it is within a safe range. The specific method is as follows:

[0070] S3.1.1. Setting the safety range of transmission line stress [F min ,F max ], F min is the minimum stress limit, F max is the maximum stress limit;

[0071] S3.1.2. Calculate the transmission line stress S(t) based on the transmission line temperature and humidity-stress influence model;

[0072] S3.1.3. Determine whether the transmission line stress S(t) is within the safe range [F min ,F max ];

[0073] If F min ≤S(t)≤F max , there is no need to modify the transmission line working parameters;

[0074] If S(t) < F min ∨F max < S(t), execute the parameter correction module;

[0075] The parameter correction module is used to correct the working parameters of the transmission line in real time when the stress of the transmission line is not within the safe range. The specific method is as follows:

[0076] When the stress S(t) of the transmission line is not within the safe range [F min , F max , then:

[0077] S3.2.1. Reduce the current load;

[0078] S3.2.2. Modify the transmission line;

[0079] S3.2.3. Manually or by a robot, cool down and dehumidify the transmission line.

[0080] As a further improvement of this technical solution, the data processing unit is used to receive the data of the data acquisition and processing unit, the environmental factor modeling unit, and the real-time analysis and correction unit, and display the wind speed, temperature, humidity, circuit load data, and the temperature and humidity - stress influence model data of the transmission line through the digital twin management platform;

[0081] When the stress of the transmission line is not within the safe range, an alarm is issued in the digital twin management platform in real time.

[0082] Compared with the prior art, the beneficial effects of the present invention are:

[0083] 1. In the AI model algorithm for a transmission line affected by natural conditions, based on the application of the time series convolutional neural network fusion model and the attention mechanism, the time series features of environmental factors such as wind speed, temperature, and humidity are accurately extracted, and through the method of dynamic fusion and weighting, the spatio-temporal correlation between environmental factors is captured more effectively.

[0084] 2. In the AI model algorithm for a transmission line affected by natural conditions, by introducing a time-varying dynamic correction factor and combining it with a non-linear regression model, the dynamic correction and real-time adjustment of the working parameters of the transmission line are realized. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Figure 1 is the overall flow block diagram of the present invention;

[0086] The meanings of the reference numerals in the figure are as follows:

[0087] 1. Data acquisition and processing unit; 2. Environmental factor modeling unit; 3. Real-time analysis and correction unit; 4. Data processing unit; 11. Data acquisition module; 12. Data processing module; 13. Data storage module; 31. Model analysis module; 32. Parameter correction module. DETAILED DESCRIPTION

[0088] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0089] See also Figure 1 As shown, an AI model algorithm for the influence of natural conditions on transmission lines is provided, including:

[0090] A data acquisition and processing unit 1, wherein the data acquisition and processing unit 1 uses sensors to collect wind speed, temperature, humidity and circuit load data in real time, uses a time series convolutional neural network fusion model to perform time series modeling on wind speed, temperature and humidity to extract their time series features, and uses an attention mechanism to fuse the wind speed, temperature and humidity time series features to obtain fusion features, and stores the fusion features in a time series database;

[0091] In this embodiment, the data acquisition and processing unit 1 includes a data acquisition module 11, a data processing module 12 and a data storage module 13;

[0092] The data acquisition module 11 uses a sensor to collect wind speed V in real time. wind (t), temperature T(t), humidity H(t) and circuit load data L(t);

[0093] Where, t is time;

[0094] The data processing module 12 uses a time series convolutional neural network fusion model to perform time series modeling on wind speed, temperature and humidity to extract their time series features, and uses an attention mechanism to fuse the time series features of wind speed, temperature and humidity to obtain fusion features;

[0095] The data storage module 13 is used to store the fusion features in a time series database;

[0096] The time series convolutional neural network fusion model is implemented based on the fusion of convolutional neural network and one-dimensional convolution, and the corresponding time series features are extracted by performing convolution operations on wind speed, temperature and humidity;

[0097] The attention mechanism is used to fuse the temporal features of wind speed, temperature and humidity.

[0098] In this embodiment, the data processing module 12 uses a time series convolutional neural network fusion model to perform time series modeling on wind speed, temperature and humidity to extract their time series features, and uses an attention mechanism to fuse the time series features of wind speed, temperature and humidity to obtain fusion features. The specific method is as follows:

[0099] S1.2.1. Use the time series convolutional neural network fusion model to model the wind speed, temperature and humidity and extract their time series characteristics:

[0100] Wind speed time series characteristics:

[0101] X wind (t) = σ(W wind ·V wind (t)+b wind );

[0102] Temperature timing characteristics:

[0103] X temp (t) = σ(W temp ·T(t)+b temp );

[0104] Humidity time series characteristics:

[0105] X humidity (t) = σ(W humidity ·H(t)+b humidity );

[0106] Among them, X wind (t) is the time series characteristic of wind speed; X temp (t) is the temperature time series characteristic; X humidity (t) is the humidity time series feature; σ is the ReLU function; W wind is the wind speed convolution kernel; b wind is the wind speed bias term; W temp is the temperature convolution kernel; b temp is the temperature bias term; W humidity is the humidity convolution kernel; b humidity is the humidity bias term; t is the time;

[0107] S1.2.2. Use the attention mechanism to calculate wind speed score, temperature score and humidity score:

[0108] Wind Speed ​​Rating:

[0109] score wind (t) = f(W′ wind ·X wind (t)+bwind );

[0110] Temperature Rating:

[0111] score temp (t) = f(W′ temp ·X temp (t)+b temp );

[0112] Moisture Rating:

[0113] score humidity (t) = f(W′ humidity ·X humidity (t)+b humidity );

[0114] Among them, score wind (t) is the wind speed score; score temp (t) is the temperature score; score humidity (t) is the humidity score; W' wind is the wind speed weight matrix; W' temp is the temperature weight matrix; W' humidity is the humidity weight matrix; f is the Sigmoid activation function;

[0115] S1.2.3. Calculate the attention weights of wind speed, temperature and humidity based on wind speed score, temperature score and humidity score:

[0116] Wind speed attention weight:

[0117]

[0118] Temperature attention weight:

[0119]

[0120] Humidity attention weight:

[0121]

[0122] Among them, α wind (t) is the wind speed attention weight; α temp (t) is the temperature attention weight; α humidity (t) is the humidity attention weight;

[0123] S1.2.4. Based on the attention weights of wind speed, temperature and humidity, the time series features of wind speed, temperature and humidity are weighted and fused to obtain the fused features:

[0124] X fused (t) = α wind (t)·Xwind (t)+α temp (t)·X temp (t)+α humidity (t)·X humidity (t);

[0125] Among them, X fused (t) is the fusion feature.

[0126] It also includes an environmental factor modeling unit 2, which uses a nonlinear regression model to establish a transmission line temperature, humidity and stress impact model based on the fusion characteristics of wind speed, temperature and humidity and introduces a time-varying dynamic correction factor;

[0127] In this embodiment, the environmental factor modeling unit 2 uses a nonlinear regression model to establish a transmission line temperature and humidity-stress influence model based on the fusion characteristics of wind speed, temperature and humidity and introduces a time-varying dynamic correction factor. The specific method steps are as follows:

[0128] S2.1. Calculate the thermal expansion of the transmission line based on temperature, calculate the wind pressure of the transmission line based on wind speed, and calculate the humidity load of the transmission line based on humidity;

[0129] S2.2, define the time-varying dynamic correction factor;

[0130] S2.3. Based on the thermal expansion of the transmission line, the wind pressure of the transmission line, the humidity load of the transmission line and the time-varying dynamic correction factor, a nonlinear regression model is used to construct the temperature, humidity and stress influence model of the transmission line.

[0131] In this embodiment S2.1, the thermal expansion of the transmission line is calculated based on the temperature, the wind pressure of the transmission line is calculated based on the wind speed, and the humidity load of the transmission line is calculated based on the humidity. The specific method is as follows:

[0132] S2.1.1. Calculation of thermal expansion of transmission lines based on temperature:

[0133] ΔL=α1L0(T(t)-T0);

[0134] Among them, ΔL is the thermal expansion of the transmission line; α1 is the linear expansion coefficient of the transmission line material; L0 is the initial length of the transmission line; T0 is the initial temperature;

[0135] S2.1.2. Calculate the wind pressure on the transmission line based on wind speed:

[0136] P wind =C d ·ρ air ·A wire ·V wind (t) 2 ;

[0137] Among them, P windis the wind pressure on the transmission line; C d is the drag coefficient; ρ air is the air density; A wire is the cross-sectional area of ​​the transmission line;

[0138] S2.1.3. Calculation of humidity load on transmission lines based on humidity

[0139] P humidity =k humidity ·A wire H(t);

[0140] Among them, P humidity is the humidity load of the transmission line; k humidity is the constant for humidity attachment.

[0141] In this embodiment S2.2, a time-varying dynamic correction factor is defined, and the specific method is as follows:

[0142]

[0143] Among them, β(t) is the time-varying dynamic correction factor; γ1 and γ2 are constants of the correction factor; λ1 is the attenuation rate factor; λ2 is the periodic factor; δ is the phase factor.

[0144] In this embodiment, the time-varying dynamic correction factor β(t) is used to introduce the dynamic effects of wind speed, temperature and humidity factors on the stress of the transmission line over time;

[0145] It is an exponential decay term, which describes the gradual weakening of the effects of thermal expansion of the transmission line over time, reflecting the behavior of natural phenomena decaying over time. When the transmission line is exposed to strong winds or high temperatures, the initial stress may be large, but as time goes by, the changes in temperature and wind speed may gradually stabilize, thereby gradually weakening the impact on the stress. To indicate such a fading influence;

[0146] γ2·sin(λ2t+δ) is a periodic term. Changes in day and night, seasonal changes, etc. will affect the temperature and wind speed of the transmission line. During the day, the temperature gradually rises, causing the transmission line to expand. At night, the temperature drops and the transmission line contracts. This change is periodic, so the periodic term γ2·sin(λ2t+δ) is used to represent the periodic influence.

[0147] In this embodiment S2.3, based on the thermal expansion of the transmission line, the wind pressure of the transmission line, the humidity load of the transmission line and the time-varying dynamic correction factor, a nonlinear regression model is used to construct a transmission line temperature, humidity and stress influence model. The specific method is as follows:

[0148] Transmission line temperature, humidity and stress impact model:

[0149] S(t) = θ0 + θ1·ΔL + θ2·P wind + θ3·P humidity + β(t)·X fused (t);

[0150] Wherein, θ0 is a constant coefficient; θ1 is the weight of the thermal expansion of the transmission line; θ2 is the wind pressure weight of the transmission line; θ3 is the humidity load weight of the transmission line; S(t) is the stress of the transmission line.

[0151] It further includes a real-time analysis and correction unit 3, and the real-time analysis and correction unit 3 dynamically adjusts the working parameters of the transmission line based on the temperature and humidity-stress influence model of the transmission line in combination with the circuit load data;

[0152] In this embodiment, the real-time analysis and correction unit 3 includes a model analysis module 31 and a parameter correction module 32;

[0153] Wherein, the model analysis module 31 calculates the stress of the transmission line based on the temperature and humidity-stress influence model of the transmission line and determines whether it is within the safe range;

[0154] The parameter correction module 32 is used to correct the working parameters of the transmission line in real time when the stress of the transmission line is not within the safe range.

[0155] In this embodiment, the model analysis module 31 calculates the stress of the transmission line based on the temperature and humidity-stress influence model of the transmission line and determines whether it is within the safe range. The specific method is as follows:

[0156] S3.1.1. Set the safe range of the transmission line stress [F min , F max , F min is the minimum stress limit, and F max is the maximum stress limit;

[0157] S3.1.2. Calculate the transmission line stress S(t) based on the temperature and humidity-stress influence model of the transmission line;

[0158] S3.1.3. Determine whether the transmission line stress S(t) is within the safe range [F min , F max ;

[0159] If F min ≤ S(t) ≤ F max , there is no need to correct the working parameters of the transmission line;

[0160] If S(t) < F min ∨ F max < S(t), execute the parameter correction module 32;

[0161] In this embodiment, the parameter correction module 32 is used to correct the transmission line working parameters in real time when the transmission line stress is not within the safe range. The specific method is as follows:

[0162] The transmission line stress S(t) is not within the safe range [F min ,F max ]hour:

[0163] S3.2.1, reduce the current load;

[0164] S3.2.2, modify the transmission line;

[0165] S3.2.3. Artificially cool and dehumidify the transmission lines.

[0166] It also includes a data processing unit 4, which displays wind speed, temperature, humidity, circuit load data and transmission line temperature and humidity-stress impact model data through the digital twin management platform;

[0167] In this embodiment, the data processing unit 4 is used to receive data from the data acquisition processing unit 1, the environmental factor modeling unit 2 and the real-time analysis and correction unit 3, and display the wind speed, temperature, humidity, circuit load data and the transmission line temperature and humidity-stress impact model data through the digital twin management platform;

[0168] When the transmission line stress is not within a safe range, an alarm is issued in real time on the digital twin management platform.

[0169] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and the above embodiments and descriptions are only preferred examples of the present invention, and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. An AI model algorithm for the influence of natural conditions on power transmission lines, characterized in that: include: A data acquisition processing unit (1), wherein the data acquisition processing unit (1) uses sensors to collect wind speed, temperature, humidity and circuit load data in real time, uses a time series convolutional neural network fusion model to perform time series modeling on the wind speed, temperature and humidity to extract their time series features, uses an attention mechanism to fuse the wind speed, temperature and humidity time series features to obtain fusion features, and stores the fusion features in a time series database; An environmental factor modeling unit (2), wherein the environmental factor modeling unit (2) establishes a transmission line temperature, humidity and stress influence model based on the fusion characteristics of wind speed, temperature and humidity and introduces a time-varying dynamic correction factor using a nonlinear regression model; A real-time analysis and correction unit (3), wherein the real-time analysis and correction unit (3) dynamically adjusts the operating parameters of the power transmission line based on a power transmission line temperature, humidity and stress influence model combined with circuit load data; A data processing unit (4), wherein the data processing unit (4) displays wind speed, temperature, humidity, circuit load data and transmission line temperature and humidity-stress impact model data through a digital twin management platform.

2. The AI ​​model algorithm for the influence of natural conditions on power transmission lines according to claim 1 is characterized in that: The data acquisition and processing unit (1) comprises a data acquisition module (11), a data processing module (12) and a data storage module (13); The data acquisition module (11) uses a sensor to collect wind speed V in real time. wind (t), temperature T(t), humidity H(t) and circuit load data L(t); Where, t is time; The data processing module (12) uses a time series convolutional neural network fusion model to perform time series modeling on wind speed, temperature and humidity to extract their time series features, and uses an attention mechanism to fuse the time series features of wind speed, temperature and humidity to obtain fusion features; The data storage module (13) is used to store the fused features in a time series database; The time series convolutional neural network fusion model is implemented based on the fusion of convolutional neural network and one-dimensional convolution, and the corresponding time series features are extracted by performing convolution operations on wind speed, temperature and humidity; The attention mechanism is used to fuse the temporal features of wind speed, temperature and humidity.

3. The AI ​​model algorithm for the influence of natural conditions on power transmission lines according to claim 2 is characterized in that: The data processing module (12) uses a time series convolutional neural network fusion model to perform time series modeling on wind speed, temperature and humidity to extract their time series features, and uses an attention mechanism to fuse the time series features of wind speed, temperature and humidity to obtain fusion features. The specific method is as follows: S1.2.

1. Use the time series convolutional neural network fusion model to model the wind speed, temperature and humidity and extract their time series characteristics: Wind speed time series characteristics: X wind (t)=σ(W wind ·V wind (t)+b wind ); Temperature timing characteristics: X temp (t)=σ(W temp ·T(t)+b temp ); Humidity time series characteristics: X humidity (t)=σ(W humidity ·H(t)+b humidity ); Among them, X wind (t) is the time series characteristic of wind speed; X temp (t) is the temperature time series characteristic; X humidity (t) is the humidity time series feature; σ is the ReLU function; W wind is the wind speed convolution kernel; b wind is the wind speed bias term; W temp is the temperature convolution kernel; b temp is the temperature bias term; W humidity is the humidity convolution kernel; b humidity is the humidity bias term; t is the time; S1.2.

2. Use the attention mechanism to calculate wind speed score, temperature score and humidity score: Wind Speed ​​Rating: score wind (t)=f(W′ wind ·X wind (t)+b wind ); Temperature Rating: score temp (t)=f(W′ temp ·X temp (t)+b temp ); Moisture Rating: score humidity (t)=f(W′ humidity ·X humidity (t)+b humidity ); Among them, score wind (t) is the wind speed score; score temp (t) is the temperature score; score humidity (t) is the humidity score; W′ wind is the wind speed weight matrix; W′ temp is the temperature weight matrix; W′ humidity is the humidity weight matrix; f is the Sigmoid activation function; S1.2.

3. Calculate the attention weights of wind speed, temperature and humidity based on wind speed score, temperature score and humidity score: Wind speed attention weight: Temperature attention weight: Humidity attention weight: Among them, α wind (t) is the wind speed attention weight; α temp (t) is the temperature attention weight; α humidity (t) is the humidity attention weight; S1.2.

4. Based on the attention weights of wind speed, temperature and humidity, the time series features of wind speed, temperature and humidity are weighted and fused to obtain the fused features: X fused (t)=α wind (t)·X wind (t)+α temp (t)·X temp (t)+α humidity (t)·X humidity (t); Among them, X fused (t) is the fusion feature.

4. The AI ​​model algorithm for the influence of natural conditions on power transmission lines according to claim 3 is characterized by: The environmental factor modeling unit (2) is based on the fusion characteristics of wind speed, temperature and humidity and introduces a time-varying dynamic correction factor, and uses a nonlinear regression model to establish a transmission line temperature, humidity and stress impact model. The specific method steps are as follows: S2.

1. Calculate the thermal expansion of the transmission line based on temperature, calculate the wind pressure of the transmission line based on wind speed, and calculate the humidity load of the transmission line based on humidity; S2.2, define the time-varying dynamic correction factor; S2.

3. Based on the thermal expansion of the transmission line, the wind pressure of the transmission line, the humidity load of the transmission line and the time-varying dynamic correction factor, a nonlinear regression model is used to construct the temperature, humidity and stress influence model of the transmission line.

5. The AI ​​model algorithm for the influence of natural conditions on power transmission lines according to claim 4 is characterized in that: In S2.1, the thermal expansion of the transmission line is calculated based on the temperature, the wind pressure of the transmission line is calculated based on the wind speed, and the humidity load of the transmission line is calculated based on the humidity. The specific method is as follows: S2.1.

1. Calculation of thermal expansion of transmission lines based on temperature: ΔL=α1L0(T(t)-T0); Among them, ΔL is the thermal expansion of the transmission line; α1 is the linear expansion coefficient of the transmission line material; L0 is the initial length of the transmission line; T0 is the initial temperature; S2.1.

2. Calculate the wind pressure on the transmission line based on wind speed: P wind =C d ·ρ air ·A wire ·V wind (t) 2 ; Among them, P wind is the wind pressure on the transmission line; C d is the drag coefficient; ρ air is the air density; A wire is the cross-sectional area of ​​the transmission line; S2.1.

3. Calculation of humidity load on transmission lines based on humidity P humidity =k humidity ·A wire ·H(t); Among them, P humidity is the humidity load of the transmission line; k humidity is the constant for humidity attachment.

6. The AI ​​model algorithm for the influence of natural conditions on power transmission lines according to claim 5 is characterized by: In S2.2, a time-varying dynamic correction factor is defined, and the specific method is as follows: Among them, β(t) is the time-varying dynamic correction factor; γ1 and γ2 are constants of the correction factor; λ1 is the attenuation rate factor; λ2 is the periodic factor; δ is the phase factor.

7. The AI ​​model algorithm for the influence of natural conditions on power transmission lines according to claim 6 is characterized by: In S2.3, based on the thermal expansion of the transmission line, the wind pressure of the transmission line, the humidity load of the transmission line and the time-varying dynamic correction factor, a nonlinear regression model is used to construct a temperature, humidity and stress influence model of the transmission line. The specific method is as follows: Transmission line temperature, humidity and stress impact model: S(t)=θ0+θ1·ΔL+θ2·P wind +θ3·P humidity +β(t)·X fused (t); Among them, θ0 is the constant coefficient; θ1 is the weight of thermal expansion of the transmission line; θ2 is the weight of wind pressure of the transmission line; θ3 is the weight of humidity load of the transmission line; S(t) is the stress of the transmission line.

8. The AI ​​model algorithm for the influence of natural conditions on power transmission lines according to claim 7 is characterized by: The real-time analysis and correction unit (3) comprises a model analysis module (31) and a parameter correction module (32); The model analysis module (31) calculates the stress of the transmission line based on the temperature, humidity and stress influence model of the transmission line, and determines whether it is within a safe range; The parameter correction module (32) is used to correct the transmission line operating parameters in real time when the transmission line stress is not within a safe range.

9. The AI ​​model algorithm for the influence of natural conditions on power transmission lines according to claim 8, characterized in that: The model analysis module (31) calculates the transmission line stress based on the transmission line temperature, humidity and stress influence model, and determines whether it is within a safe range. The specific method is as follows: S3.1.

1. Setting the safety range of transmission line stress [F min , F max ], F min is the minimum stress limit, F max is the maximum stress limit; S3.1.

2. Calculate the transmission line stress S(t) based on the transmission line temperature and humidity-stress influence model; S3.1.

3. Determine whether the transmission line stress S(t) is within the safe range [F min , F max ]; If F min ≤S(t)≤F max , there is no need to modify the transmission line working parameters; If S(t) < F min ∨ F max < S(t), execute the parameter correction module (32); The parameter correction module (32) is used to correct the transmission line working parameters in real time when the transmission line stress is not within the safe range. The specific method is as follows: The transmission line stress S(t) is not within the safe range [F min , F max ]hour: S3.2.1, reduce the current load; S3.2.2, modify the transmission line; S3.2.

3. Artificially cool and dehumidify the transmission lines.

10. The AI ​​model algorithm for the influence of natural conditions on power transmission lines according to claim 9 is characterized in that: The data processing unit (4) is used to receive data from the data acquisition processing unit (1), the environmental factor modeling unit (2) and the real-time analysis and correction unit (3), and to display wind speed, temperature, humidity, circuit load data and transmission line temperature, humidity and stress impact model data through the digital twin management platform; When the transmission line stress is not within a safe range, an alarm is issued in real time on the digital twin management platform.

Images