An image analysis-based regional power facility state analysis method and system

Abstract

The present application relates to a kind of regional power facility state analysis method based on image analysis, belong to the field of image reconstruction, more specifically belong to the field of image generation.The method comprises: using AI state analysis model based on the standard pattern of each imaging perspective of current tower and the multiple visual parameters of the erected line area of current tower Intelligent analysis of the distance between two erected lines of erected line area and the overall surface corrosion area proportion of current tower;Based on the intelligent analysis result, it is decided whether to trigger the wireless alarm operation associated with the current tower state.The present application also relates to a kind of regional power facility state analysis system based on image analysis.Through the present application, for the technical problem that the distance between two erected lines and the overall corrosion degree of tower are difficult to be synchronously intelligent analyzed, artificial intelligence mode is used to realize the above-mentioned synchronous intelligent analysis based on the basic information of the multiple visual parameters of regional power facility, thereby solving the above-mentioned technical problems.

Description

TECHNICAL FIELD

[0001] The image reconstruction of the present application belongs to the field of image generation, and particularly relates to a regional power facility state analysis method and system based on image analysis. BACKGROUND

[0002] High-voltage transmission lines are composed of different power facilities in various regions, and the state monitoring of these power facilities is crucial for the normal operation of high-voltage transmission lines. However, the special environment of high voltage and the outdoor work scene result in the on-site setting of various state monitoring facilities and multiple state monitoring personnel for real-time state monitoring of various power facilities. This not only wastes time and effort, but also is not practical. Therefore, an unmanned state monitoring mechanism is generally selected to perform real-time state monitoring on different power facilities in various regions of high-voltage transmission lines. These unmanned state monitoring mechanisms focus on the state monitoring of power facilities such as erected lines, towers, and insulators.

[0003] For example, Chinese invention patent publication CN118554636A proposes a power facility unmanned inspection method and device for a power storage station. The method includes obtaining the attributes of the power storage station power facility, determining the key parts thereof according to the attributes, and setting sensors for monitoring the operation data of the power storage station power facility at the key parts. The sensors transmit the collected operation data to a data center through the Internet of Things. The data center processes and analyzes the operation data, identifies the operation status and existing problems of the power storage station power facility through an identification model, and discovers abnormal states or faults of the power storage station power facility. When an abnormality or fault of the power storage station power facility is discovered, the data center performs operations such as restarting the equipment or adjusting the parameters through a remote control function to eliminate the abnormality or fault. The device includes a data acquisition module, a data analysis module, and a result output module. The present application realizes comprehensive management and maintenance of the equipment, and ensures the stable operation and safety of the power system.

[0004] For example, Chinese invention patent publication CN115575755A proposes an electric power facility inspection system based on artificial intelligence and Beidou technology. The system includes a line region division module, a line image and information collection module, a thermal image temperature analysis module, a temperature influence analysis module, a line temperature compliance analysis module, a line loss compliance analysis module, a line comprehensive state analysis module, a warning terminal, and a database. The system analyzes the temperature of the transmission line and the environment of the transmission line region, and then analyzes the temperature compliance and loss compliance of the transmission line. The system solves the problem of rough and general analysis of the temperature of the transmission line in the current technology, realizes multi-dimensional analysis of the temperature of the transmission line, effectively ensures the reliability and reference of the detection and analysis results of the temperature of the transmission line, and effectively ensures the safety of electric power transportation of the transmission line, as well as the efficiency and effect of the transmission line during transportation.

[0005] However, the above prior art all lack an unmanned state analysis mechanism for the overall surface corrosion degree of each tower of the high-voltage transmission line and whether the distance between each pair of lines erected by the tower meets the standard, and naturally lack a synchronous unmanned state analysis mechanism for the overall surface corrosion degree of each tower of the high-voltage transmission line and whether the distance between each pair of lines erected by the tower meets the standard, resulting in the need to use on-site measurement to measure the different states of the above two power facilities, and even so, it is still very difficult to determine whether the overall surface corrosion degree of each tower and the distance between each pair of lines erected by the tower meets the standard when performing on-site measurement. SUMMARY

[0006] To solve the technical problems in the prior art, the present application provides a regional power facility state analysis method and system based on image analysis, which performs synchronous unmanned intelligent analysis on the overall surface corrosion degree of each tower of the high-voltage transmission line and whether the distance between each pair of lines erected by the tower meets the standard based on a customized structure design AI state analysis model and a targeted selection of various basic information including multiple visual parameters of regional power facilities, thereby saving a large amount of on-site equipment and labor costs and providing reliable protection for the safe operation of high-voltage transmission lines.

[0007] According to a first aspect of the present application, a regional power facility state analysis method based on image analysis is provided, the method comprising:

[0008] performing a downward-looking picture collection action on the erected line region of the current tower of the high-voltage transmission line to obtain a regional imaging picture corresponding to the erected line region of the current tower;

[0009] taking the installation height of the current tower, the length, the number, the cross-sectional area, and the erection height of the erected line of the erected line region of the current tower as the configuration information of the erected line region of the current tower;

[0010] performing each training operation on the feedforward neural network to obtain the feedforward neural network after performing each training operation and as an AI state analysis model;

[0011] using the AI state analysis model to intelligently analyze the distance between each pair of lines erected by the current tower and the overall surface corrosion area ratio of the current tower based on the respective standard patterns of the current tower at each imaging angle, the color component numerical range corresponding to the erected line of the erected line region of the current tower, the color component numerical value of each pixel point of the regional imaging picture corresponding to the erected line region of the current tower, the depth of field numerical value and coordinate numerical value, and the configuration information of the erected line region of the current tower;

[0012] The wireless alarm device is connected with the state analysis device, and is used for determining whether to trigger a wireless alarm operation associated with the current tower state based on the intelligent analysis result.

[0013] According to a second aspect of the present application, there is provided an image analysis-based regional power facility state analysis system, comprising:

[0014] A regional acquisition device is used for performing a low-altitude aerial image acquisition operation on a tower erection line region of a current tower of a high-voltage transmission line to obtain a regional imaging image corresponding to the tower erection line region of the current tower;

[0015] A configuration analysis device is used for taking the installation height of the current tower, the length, the number, the cross-sectional area and the erection height of the tower erection line of the tower erection line region of the current tower as the configuration information of the tower erection line region of the current tower;

[0016] A sequential construction device is used for performing a plurality of training operations on the feedforward neural network to obtain a feedforward neural network after the plurality of training operations are performed and as an AI state analysis model;

[0017] A state analysis device is connected with the regional acquisition device, the configuration analysis device and the sequential construction device respectively, and is used for intelligently analyzing the tower erection line-to-tower erection line spacing of the tower erection line region of the current tower and the overall surface corrosion area proportion of the current tower based on the AI state analysis model, the respective standard patterns of the current tower at respective imaging angles, the color component numerical value range of the tower erection line of the tower erection line region of the current tower, the color component numerical value, the depth of field numerical value and the coordinate numerical value of each pixel point of the regional imaging image corresponding to the tower erection line region of the current tower, and the configuration information of the tower erection line region of the current tower;

[0018] A wireless alarm device is connected with the state analysis device, and is used for determining whether to trigger a wireless alarm operation associated with the current tower state based on the intelligent analysis result.

[0019] Therefore, the present application has at least the following five outstanding substantive features:

[0020] Substantive feature A: an image analysis-based artificial intelligence analysis mode is adopted to synchronously complete non-contact intelligent analysis of the tower erection line-to-tower erection line spacing of the tower erection line region of each tower of the high-voltage transmission line and the overall surface corrosion area proportion of each tower, and it is determined whether to perform state alarm of different power facilities based on two different state data obtained by analysis, so that synchronous intelligent analysis of the tower erection line-to-tower erection line spacing and the overall corrosion degree of the tower is realized, and the automation level and the intelligent level of the high-voltage transmission line management are improved;

[0021] Substantial feature B: in order to synchronously complete the non-contact intelligent analysis of the erection line distance between two erection lines and the overall surface corrosion area ratio of the current tower in the erection line area of the current tower of the high-voltage transmission line, an AI state analysis model of customized structure design is introduced, the AI state analysis model is a feedforward neural network after performing each training operation, and the number of times of performing training operation of the feedforward neural network is positively correlated with the total length of the high-voltage transmission line, so as to customize AI state analysis models of different structures for different high-voltage transmission lines, and ensure the effectiveness and stability of the synchronous intelligent analysis results of the erection line distance between two erection lines and the overall corrosion degree of the tower;

[0022] Substantial feature C: in order to synchronously complete the non-contact intelligent analysis of the erection line distance between two erection lines and the overall surface corrosion area ratio of the current tower in the erection line area of the current tower of the high-voltage transmission line, each item of basic data is introduced, the basic data includes each standard pattern corresponding to the current tower under each imaging angle, the color component value range corresponding to the erection line of the erection line area of the current tower, the color component value, the depth of field value and the coordinate value of each pixel point of the area imaging picture corresponding to the erection line area of the current tower, and each configuration information of the erection line area of the current tower, the targeted selection of the above basic data further ensures the effectiveness and stability of the synchronous intelligent analysis results of the erection line distance between two erection lines and the overall corrosion degree of the tower;

[0023] Substantial feature D: specifically, each standard pattern corresponding to the current tower under each imaging angle only includes the current tower and the resolution of each standard pattern is the same, which is equal to the resolution of the area imaging picture corresponding to the erection line area of the current tower, the color component value range corresponding to the erection line of the erection line area of the current tower is the red-green component value range, the black-white component value range and the yellow-blue component value range corresponding to the erection line of the erection line area of the current tower, each component value range is limited by the upper threshold value and the lower threshold value of the corresponding component value, the color component value of each pixel point is the red-green component value, the black-white component value and the yellow-blue component value of the pixel point, and the installation height of the current tower, the length, the number, the cross-sectional area and the erection height of the erection line of the erection line area of the current tower are taken as the configuration information of the erection line area of the current tower, so as to design the data structure of each input content of the AI state analysis model;

[0024] Essential feature E: in each training operation performed on the feedforward neural network, the known erection line two-two distance of the erection line area of a certain tower and the known overall surface corrosion area proportion of the certain tower are taken as two output contents of the feedforward neural network, and each corresponding standard pattern of the certain tower at each imaging angle, the color component numerical range corresponding to the erection line of the erection line area of the certain tower, the color component numerical value of each pixel point of the area imaging picture corresponding to the erection line area of the certain tower, the depth of field numerical value and coordinate numerical value, and each configuration information of the erection line area of the certain tower are taken as each output content of the feedforward neural network, and the training operation is completed, so as to ensure the training effect of each training operation of the feedforward neural network. BRIEF DESCRIPTION OF DRAWINGS

[0025] The embodiments of the application will be described below with reference to the accompanying drawings, in which:

[0026] Figure 1 It is a working scene schematic diagram of a regional power facility state analysis method and system based on image analysis according to the application.

[0027] Figure 2 It is a step flowchart of a regional power facility state analysis method based on image analysis according to embodiment 1 of the application.

[0028] Figure 3 It is a step flowchart of a regional power facility state analysis method based on image analysis according to embodiment 2 of the application.

[0029] Figure 4 It is a step flowchart of a regional power facility state analysis method based on image analysis according to embodiment 3 of the application.

[0030] Figure 5 It is an internal structure diagram of a regional power facility state analysis system based on image analysis according to embodiment 4 of the application.

[0031] Figure 6 It is an internal structure diagram of a regional power facility state analysis system based on image analysis according to embodiment 5 of the application.

[0032] Figure 7 It is an internal structure diagram of a regional power facility state analysis system based on image analysis according to embodiment 6 of the application. DETAILED DESCRIPTION

[0033] As Figure 1As shown, a working scene schematic diagram of an image analysis-based regional power facility state analysis method and system according to the present application is given, and the image reconstruction of the present application belongs to the field of image generation.

[0034] The specific technical process of the present application is as follows:

[0035] Technical process A: for intelligent prediction of the order setting time of the target user of each benefit type for ordering type goods, a customized structure design marketing effect prediction model is introduced;

[0036] Specifically, the marketing effect prediction model at least has the following customized structure design:

[0037] First: the marketing effect prediction model is a feedforward neural network after multiple learning is performed;

[0038] Second: the number of times of learning of the feedforward neural network is positively correlated with the total number of types of goods on sale of the current interconnected marketing platform, thereby constructing marketing effect prediction models of different structures for different interconnected marketing platforms;

[0039] Third: the feedforward neural network includes a single input layer, multiple hidden layers and a single output layer, the multiple hidden layers are located between the single input layer and the single output layer, and the number of layers of the hidden layers of the feedforward neural network is positively correlated with the number of registered users of the current interconnected marketing platform;

[0040] Fourth: in each learning performed on the feedforward neural network, the interval time length from the distribution time point to the order setting time of a certain benefit type is taken as the single output content of the feedforward neural network, the total number of types of goods on sale of the current interconnected marketing platform, multiple marketing associated information of a certain target customer and single benefit data corresponding to the certain benefit type are taken as multiple input contents of the feedforward neural network, and the learning is completed, thereby ensuring the learning effect of each learning of the feedforward neural network;

[0041] In this way, through the customized structure design of the above marketing effect prediction model, the stability and reliability of the intelligent prediction result are ensured;

[0042] Technical process B: for intelligent prediction of the order setting time of the target user of each benefit type for ordering type goods, each item of basic information is selected;

[0043] Specifically, the basic information includes the total number of types of goods on sale of the current interconnected marketing platform, multiple marketing associated information of a target customer and single benefit data corresponding to each benefit type, such as Figure 1 As shown, the total number of types of goods on sale of the current interconnected marketing platform is derived from the marketing management server of the current interconnected marketing platform;

[0044] Further specifically, the multiple pieces of marketing association information of the target customer are the time length of the target customer browsing the set type goods, the interval days from the order placement date of the last purchase of the set type goods to the current day, the user registration time length, the customer age, and the customer gender, the single piece of benefit data corresponding to each benefit category is the benefit category number and the specific benefit data;

[0045] Further specifically, there are various benefit categories including discount coupons, full-reduction coupons, and gifted points, different benefit categories correspond to different benefit category numbers, when the benefit category is a discount coupon, the corresponding specific benefit data is a discount percentage value, when the benefit category is a full-reduction coupon, the corresponding specific benefit data is a full-amount plus-minus deduction amount, and when the benefit category is a gifted point, the corresponding specific benefit data is a specific point value gifted, thereby completing the data structure design of the respective pieces of benefit data corresponding to various benefit categories, and providing a data basis for the intelligent prediction of the order placement time length of the target user placing an order for the set type goods corresponding to each benefit category;

[0046] In this way, through sufficient and comprehensive selection of the above-mentioned various pieces of basic information, the stability and reliability of the intelligent prediction result are further ensured;

[0047] Technical process C: using the marketing effect prediction model customized according to the technical process A to intelligently predict the order placement time length of the target user placing an order for the set type goods corresponding to each benefit category at the current interconnection marketing platform according to the various pieces of basic information selected in the technical process B;

[0048] By way of example, the current interconnection marketing platform is an e-commerce sales platform that converges various types of goods, and there are different e-commerce sales platforms in different countries and regions, mainly manifested as an e-commerce sales APP running on a user device handheld mobile terminal, and the e-commerce sales APP corresponds to the e-commerce sales platform that converges various types of goods;

[0049] Technical process D: according to the intelligent prediction result of the various pieces of order placement time length of the target user placing an order for the set type goods corresponding to various benefits in the technical process C, selecting the benefit category corresponding to the shortest order placement time length to distribute to the target customer for the target customer to purchase the set type goods on the current interconnection marketing platform;

[0050] By way of example, when the target customer purchases the set type goods of apples through the current interconnection marketing platform, there is a discount coupon of 8 times, with a discount percentage value of 80%, there are also various full-reduction coupons including buy 100 minus 25, buy 50 minus 10, and buy 25 minus 3, and there are also the current interconnection marketing platform points of buy 100 gifted 50, buy 50 gifted 25, and buy 25 gifted 10, a total of 7 benefits can be issued to the target customer;

[0051] The present application respectively performs intelligent prediction on the order placing time length of each share of the target user placing an order for apples of each benefit, and finally issues the benefit with the shortest intelligent predicted order placing time length, such as a full-reduction coupon of 50 minus 10, to the target customer for purchasing the set type goods of apples through the current interconnection marketing platform;

[0052] It can be seen that, for the target user browsing the set type goods through the current interconnection marketing platform, the marketing effect prediction model is used to traverse various benefits to respectively intelligently predict the order placing time length of each share of the target user placing an order for the set type goods, and the benefit type with the shortest intelligent predicted order placing time length is distributed as the current benefit of the set type goods to the target customer, so as to complete the regional power facility state analysis based on image analysis for the target customer, and improve the order placing speed of each target user to ensure the overall operation efficiency and operation effect of the current interconnection marketing platform.

[0053] The key points of the present application are that the data structure design of each share of benefit data corresponding to various benefit types, the directional prediction of the marketing effect of various benefits for the specific purchase scene of the target customer purchasing the set type goods through the current interconnection marketing platform, the multi-structure design of the marketing effect prediction model, and the sufficient and comprehensive selection of each item of basic information for intelligent prediction.

[0054] In the following, a kind of regional power facility state analysis method and system based on image analysis of the present application will be described in the form of embodiments.

[0055] Embodiment 1

[0056] Figure 2 A step flow chart of a kind of regional power facility state analysis method based on image analysis according to the embodiment 1 of the present application is shown.

[0057] As Figure 2 shown, the regional power facility state analysis method based on image analysis includes the following steps:

[0058] Step S201: analyze each share of benefit data corresponding to each benefit type of the set type goods matched with the current interconnection marketing platform, and each share of benefit data corresponding to each benefit type is a benefit type number and specific benefit data;

[0059] For example, when the target customer purchases the set type of goods such as apples through the current internet marketing platform, there is a discount coupon of 8 times discount, with a discount percentage value of 80%, and there are also various full-reduction coupons including buy 100 and get 25 off, buy 50 and get 10 off, and buy 25 and get 3 off, and there are also current internet marketing platform points of buy 100 and get 50, buy 50 and get 25, and buy 25 and get 10, a total of 7 benefits can be issued to the target customer;

[0060] Step S202: The customer browsing the current internet marketing platform is taken as the target customer, and the duration of the target customer browsing the set type of goods, the interval days from the order date of the last purchase of the set type of goods to the current day, the user registration duration, the customer age, and the customer gender are taken as the multiple pieces of marketing association information of the target customer;

[0061] Specifically, a plurality of information collection components can be used to collect the duration of the target customer browsing the set type of goods, the interval days from the order date of the last purchase of the set type of goods to the current day, the user registration duration, the customer age, and the customer gender, respectively;

[0062] Step S203: The feedforward neural network is learned for multiple times to obtain the feedforward neural network after the multiple times of learning and output as the marketing effect prediction model, and the number of times of learning is positively correlated with the total number of the types of goods on sale of the current internet marketing platform;

[0063] For example, the positive correlation between the number of times of learning and the total number of the types of goods on sale of the current internet marketing platform includes: when the total number of the types of goods on sale of the current internet marketing platform is 1000, the number of times of learning selected for the feedforward neural network is 2000, when the total number of the types of goods on sale of the current internet marketing platform is 1500, the number of times of learning selected for the feedforward neural network is 2500, when the total number of the types of goods on sale of the current internet marketing platform is 2000, the number of times of learning selected for the feedforward neural network is 3000, when the total number of the types of goods on sale of the current internet marketing platform is 2500, the number of times of learning selected for the feedforward neural network is 3500, and so on;

[0064] Step S204: The marketing effect prediction model is used to intelligently predict the interval duration from the distribution time to the order time of each benefit type according to the total number of the types of goods on sale of the current internet marketing platform, the multiple pieces of marketing association information of the target customer, and the single benefit data corresponding to each benefit type;

[0065] For example, the marketing effect prediction model can be used to intelligently predict the interval time corresponding to each type of benefit, for example, the interval time corresponding to the 8-fold discount coupon is 5 minutes, the interval time corresponding to the buy 50 and get 10 discount coupon is 2 minutes, the interval time corresponding to the buy 100 and get 50 current online marketing platform points is 8 minutes, and so on.

[0066] Step S205: Obtain the interval time corresponding to each type of benefit, and distribute the benefit corresponding to the shortest interval time to the target customer as the current benefit of the set type of goods;

[0067] For example, when the interval time corresponding to the 8-fold discount coupon is 5 minutes, the interval time corresponding to the buy 50 and get 10 discount coupon is 2 minutes, and the interval time corresponding to the buy 100 and get 50 current online marketing platform points is 8 minutes, the benefit corresponding to the shortest interval time is the buy 50 and get 10 discount coupon, and therefore, the buy 50 and get 10 discount coupon is distributed to the target customer as the current benefit of the set type of goods, to complete the dynamic distribution of the intelligent benefit.

[0068] The feedforward neural network includes a single input layer, a plurality of hidden layers, and a single output layer, the plurality of hidden layers are located between the single input layer and the single output layer, and the number of hidden layers of the feedforward neural network is positively correlated with the number of registered users of the current online marketing platform.

[0069] For example, the feedforward neural network includes a single input layer, a plurality of hidden layers, and a single output layer, the plurality of hidden layers are located between the single input layer and the single output layer, and the number of hidden layers of the feedforward neural network is positively correlated with the number of registered users of the current online marketing platform, which includes: when the number of registered users of the current online marketing platform is 5 million, the number of hidden layers of the corresponding feedforward neural network is 3, when the number of registered users of the current online marketing platform is 10 million, the number of hidden layers of the corresponding feedforward neural network is 5, when the number of registered users of the current online marketing platform is 20 million, the number of hidden layers of the corresponding feedforward neural network is 7, when the number of registered users of the current online marketing platform is 30 million, the number of hidden layers of the corresponding feedforward neural network is 9, and so on.

[0070] In each learning of the feedforward neural network, the interval time from the distribution time to the order time of a known type of benefit is used as the single output content of the feedforward neural network, and the total number of types of goods on sale in the current online marketing platform, the multiple marketing association information of a target customer, and the single benefit data corresponding to the type of benefit are used as the multiple input content of the feedforward neural network, to complete the learning.

[0071] The single-portion benefit data corresponding to each benefit category includes a benefit category number and specific benefit data, and the different benefit categories correspond to different benefit category numbers.

[0072] Specifically, the different benefit categories correspond to different benefit category numbers, including using different binary numerical values of the same number of bits to represent different benefit category numbers.

[0073] The single-portion benefit data corresponding to each benefit category includes a benefit category number and specific benefit data, and the different benefit categories correspond to different benefit category numbers.

[0074] And the marketing effect prediction model intelligently predicts the interval length from the distribution time to the order time of the benefit category corresponding to each benefit category according to the total number of types of goods on sale of the current interconnection marketing platform, the multiple marketing association information of the target customer, and the single-portion benefit data corresponding to each benefit category, including intelligently predicting the interval length from the distribution time to the order time of the benefit category corresponding to each benefit category as the interval length between the time when the target customer is allocated the distribution benefit of the benefit category as a set type of goods to the target customer and the order time when the target customer orders to purchase the set type of goods using the benefit category.

[0075] Embodiment 2

[0076] Figure 3 A step flowchart of a regional power facility state analysis method based on image analysis according to the embodiment 2 of the present application is shown.

[0077] As shown in Figure 3 After obtaining the interval length corresponding to each benefit category respectively, and allocating the benefit category corresponding to the shortest interval length to the target customer as the current benefit of the set type of goods, that is, after step S205, the regional power facility state analysis method based on image analysis further includes:

[0078] Step S206: sending the benefit category corresponding to the shortest interval length to the marketing management server of the current interconnection marketing platform through the network as the reference priority benefit category of the target customer for the set type of goods;

[0079] Exemplarily, sending the benefit type corresponding to the shortest interval duration as the reference priority benefit type of the target customer for the set type commodity to the marketing management server of the current interconnection marketing platform through the network comprises: the marketing management server of the current interconnection marketing platform is a blockchain management network element, a big data management network element or a cloud computing management network element;

[0080] Wherein, sending the benefit type corresponding to the shortest interval duration as the reference priority benefit type of the target customer for the set type commodity to the marketing management server of the current interconnection marketing platform through the network comprises: the marketing management server of the current interconnection marketing platform stores different reference priority benefit types of each target customer for different types of commodities respectively using different physical addresses.

[0081] Embodiment 3

[0082] Figure 4 A step flow chart of a regional power facility state analysis method based on image analysis according to the embodiment 3 of the present application is shown.

[0083] As Figure 4 shown, before the target customer's duration of browsing the set type commodity, the interval days from the order date of the target customer's last purchase of the set type commodity to the current day, the user registration duration, the customer age and the customer gender as the target customer's multiple marketing association information are browsed, i.e. before step S202, the regional power facility state analysis method based on image analysis further comprises:

[0084] Step S207: retrieving the marketing management server of the current interconnection marketing platform to obtain the interval days from the order date of the target customer's last purchase of the set type commodity to the current day, and retrieving the user management server of the current interconnection marketing platform to obtain the target customer's duration of browsing the set type commodity, the user registration duration, the customer age and the customer gender;

[0085] Exemplarily, retrieving the marketing management server of the current interconnection marketing platform to obtain the interval days from the order date of the target customer's last purchase of the set type commodity to the current day, and retrieving the user management server of the current interconnection marketing platform to obtain the target customer's duration of browsing the set type commodity, the user registration duration, the customer age and the customer gender comprises: different numerical gender identifiers can be used to represent the customer gender of the target customer of different genders respectively.

[0086] Wherein, in the user management server of the current interconnection marketing platform, different customer management documents are provided for different target customers, and different log documents are also provided for different target customers.

[0087] and within any of the above embodiments 1-3, optionally, in the image analysis based regional power facility status analysis method:

[0088] The marketing effect prediction model is used to intelligently predict the interval time length from the distribution time to the order time corresponding to each benefit category according to the total number of types of goods for sale of the current interconnected marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category, which further comprises: performing binary value conversion processing on the total number of types of goods for sale of the current interconnected marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category, respectively, to obtain the total number of types of goods for sale of the current interconnected marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category after performing binary value conversion processing, respectively.

[0089] Specifically, performing binary value conversion processing on the total number of types of goods for sale of the current interconnected marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category, respectively, to obtain the total number of types of goods for sale of the current interconnected marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category after performing binary value conversion processing, respectively, comprises: the contents that are already in binary value representation form do not need to perform binary value conversion processing again.

[0090] The marketing effect prediction model is used to intelligently predict the interval time length from the distribution time to the order time corresponding to each benefit category according to the total number of types of goods for sale of the current interconnected marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category, which further comprises: performing binary value conversion processing on the total number of types of goods for sale of the current interconnected marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category, respectively, to obtain the total number of types of goods for sale of the current interconnected marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category after performing binary value conversion processing, respectively, to obtain the total number of types of goods for sale of the current interconnected marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category after performing binary value conversion processing, respectively.

[0091] The marketing effect prediction model is used to intelligently predict the interval time length from the distribution time to the order time corresponding to each benefit category according to the total number of types of goods for sale of the current interconnected marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category, which further comprises: the interval time length from the distribution time to the order time corresponding to each benefit category output by the marketing effect prediction model is in the form of a binary value.

[0092] The total number of types of goods for sale of the current interconnection marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category are respectively subjected to binary value conversion processing to obtain the total number of types of goods for sale of the current interconnection marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category after the binary value conversion processing.

[0093] The total number of types of goods for sale of the current interconnection marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category after the binary value conversion processing are input in parallel to the marketing effect prediction model, and the marketing effect prediction model is run to obtain the interval time length from the distribution time point to the order placement time point corresponding to the benefit category output by the marketing effect prediction model.

[0094] For example, the ASIC chip is connected with the CPLD chip.

[0095] Embodiment 4

[0096] Figure 5 An internal structure diagram of an image analysis-based regional power facility state analysis system according to Embodiment 4 of the present application is shown.

[0097] As shown in Figure 5 The image analysis-based regional power facility state analysis system includes the following components:

[0098] The benefit traversal device is configured to analyze each piece of benefit data corresponding to each benefit category matching the set type of goods of the current interconnection marketing platform, and the single piece of benefit data corresponding to each benefit category is a benefit category number and specific benefit data.

[0099] For example, when the target customer purchases the set type of goods such as apples through the current Internet marketing platform, there is a discount coupon of 8 times discount, with a discount percentage value of 80%, there are also various full price reduction coupons including buy 100 and get 25 off, buy 50 and get 10 off, and buy 25 and get 3 off, and there are also current Internet marketing platform points of buy 100 and get 50, buy 50 and get 25, and buy 25 and get 10, a total of 7 benefits can be issued to the target customer;

[0100] The customer analysis device is used to target customers who browse the current Internet marketing platform, and the time length of the target customers browsing the set type of goods, the interval days from the order date of the last purchase of the set type of goods to the current day, the user registration time length, the customer age, and the customer gender are used as multiple marketing association information of the target customers;

[0101] Specifically, a plurality of information collection components can be used to collect the time length of the target customer browsing the set type of goods, the interval days from the order date of the last purchase of the set type of goods to the current day, the user registration time length, the customer age, and the customer gender;

[0102] The multi-layer construction device is used to perform multiple learning on the feedforward neural network to obtain the feedforward neural network after performing the multiple learning and output as a marketing effect prediction model, and the number of learning is positively correlated with the total number of the set type of goods on sale of the current Internet marketing platform;

[0103] For example, the number of learning is positively correlated with the total number of the set type of goods on sale of the current Internet marketing platform, including: when the total number of the set type of goods on sale of the current Internet marketing platform is 1000, the number of learning performed on the feedforward neural network is selected to be 2000, when the total number of the set type of goods on sale of the current Internet marketing platform is 1500, the number of learning performed on the feedforward neural network is selected to be 2500, when the total number of the set type of goods on sale of the current Internet marketing platform is 2000, the number of learning performed on the feedforward neural network is selected to be 3000, when the total number of the set type of goods on sale of the current Internet marketing platform is 2500, the number of learning performed on the feedforward neural network is selected to be 3500, and so on;

[0104] The effect prediction device is connected with the benefit traversal device, the customer analysis device, and the multi-layer construction device, respectively, and is used to intelligently predict the interval time length from the distribution time to the order time of each benefit category by using the marketing effect prediction model according to the total number of the set type of goods on sale of the current Internet marketing platform, the multiple marketing association information of the target customer, and the single benefit data corresponding to each benefit category;

[0105] For example, the marketing effect prediction model can be used to intelligently predict the interval time corresponding to each type of benefit, for example, the interval time corresponding to the 8-fold discount coupon is 5 minutes, the interval time corresponding to the buy 50 and get 10 discount coupon is 2 minutes, the interval time corresponding to the buy 100 and get 50 current interconnection marketing platform points is 8 minutes, and so on.

[0106] The dynamic allocation device is connected with the effect prediction device, and is used to obtain the interval time corresponding to each type of benefit, and allocate the benefit corresponding to the shortest interval time to the target customer as the current benefit of the set type of goods.

[0107] For example, when the interval time corresponding to the 8-fold discount coupon is 5 minutes, the interval time corresponding to the buy 50 and get 10 discount coupon is 2 minutes, and the interval time corresponding to the buy 100 and get 50 current interconnection marketing platform points is 8 minutes, the benefit corresponding to the shortest interval time is the buy 50 and get 10 discount coupon, and therefore, the buy 50 and get 10 discount coupon is allocated to the target customer as the current benefit of the set type of goods, so as to complete the dynamic allocation of the intelligent benefit.

[0108] The feedforward neural network includes a single input layer, a plurality of hidden layers and a single output layer, the plurality of hidden layers are located between the single input layer and the single output layer, and the number of hidden layers of the feedforward neural network is positively correlated with the number of registered users of the current interconnection marketing platform.

[0109] For example, the feedforward neural network includes a single input layer, a plurality of hidden layers and a single output layer, the plurality of hidden layers are located between the single input layer and the single output layer, and the number of hidden layers of the feedforward neural network is positively correlated with the number of registered users of the current interconnection marketing platform, which includes that when the number of registered users of the current interconnection marketing platform is 5 million, the number of hidden layers of the corresponding feedforward neural network is 3, when the number of registered users of the current interconnection marketing platform is 10 million, the number of hidden layers of the corresponding feedforward neural network is 5, when the number of registered users of the current interconnection marketing platform is 20 million, the number of hidden layers of the corresponding feedforward neural network is 7, when the number of registered users of the current interconnection marketing platform is 30 million, the number of hidden layers of the corresponding feedforward neural network is 9, and so on.

[0110] In each learning of the feedforward neural network, the interval time from the allocation time to the order time of a certain type of benefit is used as the single output content of the feedforward neural network, and the total number of types of goods on sale of the current interconnection marketing platform, the multiple marketing association information of a certain target customer and the single benefit data corresponding to the certain type of benefit are used as the multiple input content of the feedforward neural network, so as to complete the learning.

[0111] wherein the single-portion benefit data corresponding to each benefit category is a benefit category number and specific benefit data, and the specific benefit data includes: there are various benefit categories including discount coupons, full-reduction coupons and gift points, and different benefit categories correspond to different benefit category numbers;

[0112] Specifically, the specific benefit data includes: there are various benefit categories including discount coupons, full-reduction coupons and gift points, and different benefit categories correspond to different benefit category numbers, and different benefit category numbers are represented by different binary values with the same number of bits.

[0113] wherein the single-portion benefit data corresponding to each benefit category is a benefit category number and specific benefit data, and the specific benefit data further includes: when the benefit category is a discount coupon, the corresponding specific benefit data is a discount percentage value, when the benefit category is a full-reduction coupon, the corresponding specific benefit data is a full-reduction amount, and when the benefit category is a gift point, the corresponding specific benefit data is a specific gift point value;

[0114] and wherein the interval length from the distribution time to the order time corresponding to each benefit category is intelligently predicted by the marketing effect prediction model according to the total number of types of goods on sale of the current interconnected marketing platform, the multiple portions of marketing-related information of the target customer, and the single-portion benefit data corresponding to each benefit category, and the interval length from the distribution time to the order time corresponding to each benefit category is intelligently predicted as the interval length between the time when the target customer is allocated the distribution benefit of the benefit category as a set type of goods and the order time when the target customer places an order to purchase the set type of goods.

[0115] Embodiment 5

[0116] Figure 6 An internal structure diagram of a regional power facility state analysis system based on image analysis according to Embodiment 5 of the present application is shown.

[0117] As shown in Figure 6 , the regional power facility state analysis system based on image analysis further includes:

[0118] A network transmission device connected to the dynamic allocation device, for sending the benefit category corresponding to the shortest interval length as the reference priority benefit category of the target customer for the set type of goods to the marketing management server of the current interconnected marketing platform through the network;

[0119] For example, sending the benefit category corresponding to the shortest interval length as the reference priority benefit category of the target customer for the set type of goods to the marketing management server of the current interconnected marketing platform through the network includes: the marketing management server of the current interconnected marketing platform is a blockchain management network element, a big data management network element or a cloud computing management network element;

[0120] The method further includes: sending, by the marketing management server of the current interconnection marketing platform, the reference priority benefit type corresponding to the shortest interval duration to the target customer as a reference priority benefit type of the target customer for the set type commodity through a network.

[0121] Embodiment 6

[0122] Figure 7 An internal structure diagram of an image analysis-based regional power facility state analysis system according to Embodiment 6 of the present application is shown.

[0123] As shown in Figure 7 The image analysis-based regional power facility state analysis system further includes:

[0124] The information retrieval device is connected to the customer analysis device and is configured to retrieve the marketing management server of the current interconnection marketing platform to obtain the interval days from the order date of the target customer's last purchase of the set type commodity to the current day, and retrieve the user management server of the current interconnection marketing platform to obtain the target customer's browsing duration of the set type commodity, the user registration duration, the customer age, and the customer gender.

[0125] For example, the marketing management server of the current interconnection marketing platform is retrieved to obtain the interval days from the order date of the target customer's last purchase of the set type commodity to the current day, and the user management server of the current interconnection marketing platform is retrieved to obtain the target customer's browsing duration of the set type commodity, the user registration duration, the customer age, and the customer gender, which includes that different numerical gender identifiers can be used to represent the customer gender of the target customer of different genders.

[0126] In the user management server of the current interconnection marketing platform, different customer management documents are provided for different target customers, and different log documents are also provided for different target customers.

[0127] In addition, in any of the above Embodiments 4-6, optionally, in the image analysis-based regional power facility state analysis system:

[0128] The marketing effect prediction model is used to intelligently predict the interval time length from the distribution time to the order time corresponding to each benefit type according to the total number of types of goods for sale of the current interconnection marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit type, and the total number of types of goods for sale of the current interconnection marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit type are respectively subjected to binary value conversion processing to obtain the total number of types of goods for sale of the current interconnection marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit type after the binary value conversion processing is completed;

[0129] Specifically, the total number of types of goods for sale of the current interconnection marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit type are respectively subjected to binary value conversion processing to obtain the total number of types of goods for sale of the current interconnection marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit type after the binary value conversion processing is completed, including that the content is already in a binary value representation form and does not need to be subjected to the binary value conversion processing again.

[0130] The marketing effect prediction model is used to intelligently predict the interval time length from the distribution time to the order time corresponding to each benefit type according to the total number of types of goods for sale of the current interconnection marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit type, and the total number of types of goods for sale of the current interconnection marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit type are respectively subjected to binary value conversion processing to obtain the total number of types of goods for sale of the current interconnection marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit type after the binary value conversion processing is completed, including that the content is already in a binary value representation form and does not need to be subjected to the binary value conversion processing again.

[0131] The marketing effect prediction model is used to intelligently predict the interval time length from the distribution time to the order time corresponding to each benefit type according to the total number of types of goods for sale of the current interconnection marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit type, and the total number of types of goods for sale of the current interconnection marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit type are respectively subjected to binary value conversion processing to obtain the total number of types of goods for sale of the current interconnection marketing platform, the multiple pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit type after the binary value conversion processing is completed, including that the content is already in a binary value representation form and does not need to be subjected to the binary value conversion processing again.

[0132] The total number of types of goods for sale of the current interconnection marketing platform, the plurality of pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category are respectively subjected to binary value conversion processing to obtain the total number of types of goods for sale of the current interconnection marketing platform, the plurality of pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category after the binary value conversion processing, which comprises: using a CPLD chip to respectively subject the total number of types of goods for sale of the current interconnection marketing platform, the plurality of pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category to binary value conversion processing, wherein the CPLD chip is logically designed by using a VHDL language.

[0133] The total number of types of goods for sale of the current interconnection marketing platform, the plurality of pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category after the binary value conversion processing are input in parallel to the marketing effect prediction model, and the marketing effect prediction model is run to obtain the interval time length from the distribution time point to the order placement time point corresponding to the benefit category output by the marketing effect prediction model, which comprises: using an ASIC chip to input the total number of types of goods for sale of the current interconnection marketing platform, the plurality of pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category after the binary value conversion processing in parallel to the marketing effect prediction model, wherein the ASIC chip is connected with the CPLD chip.

[0134] For example, the ASIC chip is connected with the CPLD chip to input the total number of types of goods for sale of the current interconnection marketing platform, the plurality of pieces of marketing association information of the target customer, and the single piece of benefit data corresponding to each benefit category after the binary value conversion processing in parallel to the marketing effect prediction model, wherein the ASIC chip and the CPLD chip use the same parameter configuration interface.

[0135] In addition, in the image analysis-based regional power facility state analysis method and system according to the present application:

[0136] The current tower erection line area of the high-voltage transmission line is subjected to a downward-looking picture collection action to obtain a regional imaging picture corresponding to the current tower erection line area, which comprises: using a UAV aerial photography mode to perform a downward-looking picture collection action on the current tower erection line area of the high-voltage transmission line to obtain a regional imaging picture corresponding to the current tower erection line area.

[0137] The unmanned aerial vehicle in the unmanned aerial vehicle aerial photography mode comprises an aerial photography support, an aerial photography camera, a positioning device, a height measuring instrument, a flight controller and a wireless transceiver interface.

[0138] For example, the unmanned aerial vehicle in the unmanned aerial vehicle aerial photography mode comprises an aerial photography support, an aerial photography camera, a positioning device, a height measuring instrument, a flight controller and a wireless transceiver interface, wherein the positioning device is a Beidou star positioning device or a Galileo positioning device.

[0139] The aerial photography camera is located below the aerial photography support, the aerial photography support is located at the bottom of the unmanned aerial vehicle, and the wireless transceiver interface is connected with the aerial photography camera, the positioning device, the height measuring instrument and the flight controller respectively.

[0140] Although the present application has been disclosed in various preferred embodiments as above, the present application should not be construed as merely covering the above embodiments, which are only exemplary embodiments. Any person skilled in the art can make some changes and improvements without departing from the spirit and scope of the present application, and therefore the protection scope of the present application should be defined by the claims and their equivalents.

Claims

1. A method for analyzing the status of regional power facilities based on image analysis, characterized in that, The method includes: A top-down image capture operation is performed on the current tower erection area of ​​the high-voltage transmission line to obtain an image of the area corresponding to the current tower erection area. The current tower's installation height, the length, number, cross-sectional area, and height of the overhead lines in the current tower's overhead line area are used as the configuration information for the current tower's overhead line area. Perform each training operation on the feedforward neural network to obtain the feedforward neural network after each training operation and use it as an AI state analysis model; The AI ​​state analysis model is used to intelligently analyze the spacing between the erected lines and the overall surface corrosion area of ​​the tower based on the standard patterns corresponding to the tower from various imaging perspectives, the color component range of the erected lines in the tower's erection area, the color component values, depth values, and coordinate values ​​of each pixel in the imaging image of the tower's erection area, as well as the configuration information of the tower's erection area. The system determines whether to trigger a wireless alarm operation associated with the current tower status based on the results of intelligent analysis.

2. The regional power facility status analysis method based on image analysis as described in claim 1, characterized in that: Performing training operations on the feedforward neural network to obtain the feedforward neural network after each training operation and using it as an AI state analysis model includes: the number of training operations performed on the feedforward neural network is positively correlated with the total length of the high-voltage transmission line; In each training operation performed on the feedforward neural network, the known spacing between each pair of the overhead lines in the overhead line area of ​​a certain iron tower and the known percentage of the overall surface corrosion area of ​​the iron tower are used as two output contents of the feedforward neural network. The standard patterns corresponding to the iron tower under each imaging viewpoint, the color component value range corresponding to the overhead lines in the overhead line area of ​​the iron tower, the color component value, depth value and coordinate value of each pixel in the regional imaging image corresponding to the overhead line area of ​​the iron tower, and the various configuration information of the overhead line area of ​​the iron tower are used as various output contents of the feedforward neural network to complete this training operation. Among them, each standard pattern corresponding to the current tower under each imaging angle only includes the current tower and the resolution of each standard pattern is the same, which is equal to the resolution of the area image corresponding to the erection line area of ​​the current tower. Among them, the arithmetic mean of the pairwise spacing of each pair of erected lines in the current tower erected line area is the pairwise spacing of erected lines in the current tower erected line area, and the proportion of the surface corrosion area of ​​the current tower to the total surface area of ​​the current tower is the proportion of the total surface corrosion area of ​​the current tower. Among them, the decision on whether to trigger a wireless alarm operation related to the current tower status based on the intelligent analysis results includes: when the spacing between any two lines in the current tower's erection area is not within the set spacing value range, a wireless alarm operation for exceeding the line spacing limit related to the current tower is executed; when the overall surface corrosion area of ​​the current tower accounts for more than the preset ratio limit, a wireless alarm operation for excessive surface corrosion related to the current tower is executed.

3. The regional power facility status analysis method based on image analysis as described in claim 2, characterized in that, After employing an AI state analysis model to intelligently analyze the spacing between each pair of erected lines and the overall surface corrosion area ratio of the current tower based on the standard patterns corresponding to each imaging view of the current tower under various imaging angles, the color component value range of the erected lines in the current tower's erected line area, the color component value, depth value, and coordinate value of each pixel in the imaging image of the current tower's erected line area, and various configuration information of the current tower's erected line area, the method further includes: Receive the spacing between each pair of erected lines in the current tower's erected line area and the percentage of the overall surface corrosion area of ​​the current tower, and perform a synchronous display action on the spacing between each pair of erected lines in the current tower's erected line area and the percentage of the overall surface corrosion area of ​​the current tower.

4. The regional power facility status analysis method based on image analysis as described in claim 2, characterized in that, Before using the current tower's installation height, the length, number, cross-sectional area, and height of the overhead lines in the current tower's overhead line area as configuration information for the current tower's overhead line area, the method further includes: Different physical measuring devices were used to measure the installation height of the current tower, the length of the overhead lines in the current tower's overhead line area, the number of overhead lines in the current tower's overhead line area, the cross-sectional area of ​​the overhead lines in the current tower's overhead line area, and the erection height of the overhead lines in the current tower's overhead line area.

5. The regional power facility status analysis method based on image analysis as described in any one of claims 2-4, characterized in that: The color component value range corresponding to the erection line area of ​​the current tower is the red-green component value range, black-and-white component value range, and yellow-and-blue component value range corresponding to the erection line area of ​​the current tower. Each component value range is limited by the corresponding upper limit threshold and lower limit threshold of the component value. The color component value of each pixel is the red-green component value, black-and-white component value, and yellow-and-blue component value of the pixel, as well as the L component value (red-and-green component value), A component value (black-and-white component value), and B component value (yellow-and-blue component value) of the pixel in the LAB color space. Among them, the color component value range corresponding to the erection line area of ​​the current tower is the red-green component value range, black-and-white component value range, and yellow-and-blue component value range corresponding to the erection line area of ​​the current tower. Each component value range is limited by the corresponding upper limit threshold and lower limit threshold of the component value. The color component value of each pixel is the red-green component value, black-and-white component value, and yellow-and-blue component value of the pixel. The value range of each component value of the red-green component value, black-and-white component value, and yellow-and-blue component value of the pixel is between 0 and 255. The positive correlation between the number of training operations performed by the feedforward neural network and the total length of the high-voltage transmission line includes: using a number mapping function to represent the positive correlation between the number of training operations performed by the feedforward neural network and the total length of the high-voltage transmission line; The method of using a number mapping function to represent the positive correlation between the number of training operations performed by the feedforward neural network and the total length of the high-voltage transmission line includes: in the number mapping function, the total length of the high-voltage transmission line is the input data of the number mapping function, and the number of training operations positively correlated with the total length of the high-voltage transmission line is the output data of the number mapping function; The AI ​​state analysis model intelligently analyzes the spacing between the erected lines and the overall surface corrosion area ratio of the current tower based on the standard patterns corresponding to the current tower from various imaging perspectives, the color component value range of the erected lines in the current tower's erection area, the color component value, depth value, and coordinate value of each pixel in the regional imaging image corresponding to the current tower's erection area, and various configuration information of the current tower's erection area. This includes simultaneously inputting the standard patterns corresponding to the current tower from various imaging perspectives, the color component value range of the erected lines in the current tower's erection area, the color component value, depth value, and coordinate value of each pixel in the regional imaging image corresponding to the current tower's erection area, and various configuration information of the current tower's erection area into the AI ​​state analysis model. The AI ​​state analysis model intelligently analyzes the spacing between the erected lines in the current tower's erection area and the overall surface corrosion area ratio of the current tower based on the standard patterns corresponding to each imaging view of the current tower, the color component value range of the erected lines in the current tower's erection area, the color component value, depth value, and coordinate value of each pixel in the regional imaging image corresponding to the erected lines in the current tower's erection area, and various configuration information of the erected lines in the current tower's erection area. It also includes: running the AI ​​state analysis model to obtain the spacing between the erected lines in the current tower's erection area and the overall surface corrosion area ratio of the current tower output by the AI ​​state analysis model. The process of synchronously inputting the standard patterns corresponding to the current tower from various imaging perspectives, the color component value range corresponding to the erection line of the current tower, the color component value, depth value, and coordinate value of each pixel in the regional imaging image corresponding to the erection line of the current tower, and various configuration information of the erection line area of ​​the current tower into the AI ​​state analysis model includes: performing octal value conversion on each standard pattern corresponding to the current tower from various imaging perspectives, the color component value range corresponding to the erection line of the current tower, the color component value, depth value, and coordinate value of each pixel in the regional imaging image corresponding to the erection line area of ​​the current tower, and various configuration information of the erection line area of ​​the current tower before synchronously inputting them into the AI ​​state analysis model. Among them, the AI ​​state analysis model is used to obtain the spacing between each pair of erected lines in the current tower's erected line area and the percentage of the overall surface corrosion area of ​​the current tower. Both the spacing between each pair of erected lines in the current tower's erected line area and the percentage of the overall surface corrosion area of ​​the current tower are represented in octal numerical form.

6. A regional power facility status analysis system based on image analysis, characterized in that, The system includes: The area acquisition device is used to perform overhead image acquisition of the current tower erection area of ​​the high-voltage transmission line to obtain the area imaging image corresponding to the current tower erection area. Configure the parsing device to use the current tower's installation height, the length, number, cross-sectional area, and height of the erected lines in the current tower's erected line area as various configuration information for the current tower's erected line area; The device is constructed sequentially to perform training operations on the feedforward neural network to obtain the feedforward neural network after each training operation and to serve as an AI state analysis model. The state analysis device is connected to the area acquisition device, configuration analysis device, and successive construction device, respectively. It is used to intelligently analyze the spacing between the erected lines in the current tower and the overall surface corrosion area ratio of the current tower based on the standard patterns corresponding to each imaging view of the current tower, the color component value range of the erected lines in the current tower's erected line area, the color component value, depth value, and coordinate value of each pixel in the regional imaging image corresponding to the current tower's erected line area, and various configuration information of the current tower's erected line area. The wireless alarm device, connected to the status analysis device, is used to determine whether to trigger a wireless alarm operation associated with the current tower status based on the results of intelligent analysis.

7. The regional power facility status analysis system based on image analysis as described in claim 6, characterized in that: Performing training operations on the feedforward neural network to obtain the feedforward neural network after each training operation and using it as an AI state analysis model includes: the number of training operations performed on the feedforward neural network is positively correlated with the total length of the high-voltage transmission line; In each training operation performed on the feedforward neural network, the known spacing between each pair of the overhead lines in the overhead line area of ​​a certain iron tower and the known percentage of the overall surface corrosion area of ​​the iron tower are used as two output contents of the feedforward neural network. The standard patterns corresponding to the iron tower under each imaging viewpoint, the color component value range corresponding to the overhead lines in the overhead line area of ​​the iron tower, the color component value, depth value and coordinate value of each pixel in the regional imaging image corresponding to the overhead line area of ​​the iron tower, and the various configuration information of the overhead line area of ​​the iron tower are used as various output contents of the feedforward neural network to complete this training operation. Among them, each standard pattern corresponding to the current tower under each imaging angle only includes the current tower and the resolution of each standard pattern is the same, which is equal to the resolution of the area image corresponding to the erection line area of ​​the current tower. Among them, the arithmetic mean of the pairwise spacing of each pair of erected lines in the current tower erected line area is the pairwise spacing of erected lines in the current tower erected line area, and the proportion of the surface corrosion area of ​​the current tower to the total surface area of ​​the current tower is the proportion of the total surface corrosion area of ​​the current tower. Among them, the decision on whether to trigger a wireless alarm operation related to the current tower status based on the intelligent analysis results includes: when the spacing between any two lines in the current tower's erection area is not within the set spacing value range, a wireless alarm operation for exceeding the line spacing limit related to the current tower is executed; when the overall surface corrosion area of ​​the current tower accounts for more than the preset ratio limit, a wireless alarm operation for excessive surface corrosion related to the current tower is executed.

8. The regional power facility status analysis system based on image analysis as described in claim 7, characterized in that, The system also includes: The synchronous display device, connected to the status analysis device, is used to receive the pairwise spacing between the erected lines in the current tower's erected line area and the percentage of the overall surface corrosion area of ​​the current tower, and to perform synchronous display actions on the pairwise spacing between the erected lines in the current tower's erected line area and the percentage of the overall surface corrosion area of ​​the current tower.

9. The regional power facility status analysis system based on image analysis as described in claim 7, characterized in that, The system also includes: The information measurement device, connected to the configuration analysis device, is used to measure the installation height of the current tower, the length of the overhead lines in the current tower's overhead line area, the number of overhead lines in the current tower's overhead line area, the cross-sectional area of ​​the overhead lines in the current tower's overhead line area, and the installation height of the overhead lines in the current tower's overhead line area using different physical measurement devices.

10. The regional power facility status analysis system based on image analysis as described in any one of claims 7-9, characterized in that: The color component value range corresponding to the erection line area of ​​the current tower is the red-green component value range, black-and-white component value range, and yellow-and-blue component value range corresponding to the erection line area of ​​the current tower. Each component value range is limited by the corresponding upper limit threshold and lower limit threshold of the component value. The color component value of each pixel is the red-green component value, black-and-white component value, and yellow-and-blue component value of the pixel, as well as the L component value (red-and-green component value), A component value (black-and-white component value), and B component value (yellow-and-blue component value) of the pixel in the LAB color space. Among them, the color component value range corresponding to the erection line area of ​​the current tower is the red-green component value range, black-and-white component value range, and yellow-and-blue component value range corresponding to the erection line area of ​​the current tower. Each component value range is limited by the corresponding upper limit threshold and lower limit threshold of the component value. The color component value of each pixel is the red-green component value, black-and-white component value, and yellow-and-blue component value of the pixel. The value range of each component value of the red-green component value, black-and-white component value, and yellow-and-blue component value of the pixel is between 0 and 255. The positive correlation between the number of training operations performed by the feedforward neural network and the total length of the high-voltage transmission line includes: using a number mapping function to represent the positive correlation between the number of training operations performed by the feedforward neural network and the total length of the high-voltage transmission line; The method of using a number mapping function to represent the positive correlation between the number of training operations performed by the feedforward neural network and the total length of the high-voltage transmission line includes: in the number mapping function, the total length of the high-voltage transmission line is the input data of the number mapping function, and the number of training operations positively correlated with the total length of the high-voltage transmission line is the output data of the number mapping function; The AI ​​state analysis model intelligently analyzes the spacing between the erected lines and the overall surface corrosion area ratio of the current tower based on the standard patterns corresponding to the current tower from various imaging perspectives, the color component value range of the erected lines in the current tower's erection area, the color component value, depth value, and coordinate value of each pixel in the regional imaging image corresponding to the current tower's erection area, and various configuration information of the current tower's erection area. This includes simultaneously inputting the standard patterns corresponding to the current tower from various imaging perspectives, the color component value range of the erected lines in the current tower's erection area, the color component value, depth value, and coordinate value of each pixel in the regional imaging image corresponding to the current tower's erection area, and various configuration information of the current tower's erection area into the AI ​​state analysis model. The AI ​​state analysis model intelligently analyzes the spacing between the erected lines in the current tower's erection area and the overall surface corrosion area ratio of the current tower based on the standard patterns corresponding to each imaging view of the current tower, the color component value range of the erected lines in the current tower's erection area, the color component value, depth value, and coordinate value of each pixel in the regional imaging image corresponding to the erected lines in the current tower's erection area, and various configuration information of the erected lines in the current tower's erection area. It also includes: running the AI ​​state analysis model to obtain the spacing between the erected lines in the current tower's erection area and the overall surface corrosion area ratio of the current tower output by the AI ​​state analysis model. The process of synchronously inputting the standard patterns corresponding to the current tower from various imaging perspectives, the color component value range corresponding to the erection line of the current tower, the color component value, depth value, and coordinate value of each pixel in the regional imaging image corresponding to the erection line of the current tower, and various configuration information of the erection line area of ​​the current tower into the AI ​​state analysis model includes: performing octal value conversion on each standard pattern corresponding to the current tower from various imaging perspectives, the color component value range corresponding to the erection line of the current tower, the color component value, depth value, and coordinate value of each pixel in the regional imaging image corresponding to the erection line area of ​​the current tower, and various configuration information of the erection line area of ​​the current tower before synchronously inputting them into the AI ​​state analysis model. Among them, the AI ​​state analysis model is used to obtain the spacing between each pair of erected lines in the current tower's erected line area and the percentage of the overall surface corrosion area of ​​the current tower. Both the spacing between each pair of erected lines in the current tower's erected line area and the percentage of the overall surface corrosion area of ​​the current tower are represented in octal numerical form.

Images