A method for processing multi-sampling track inspection data of electric power inspection vehicle

Abstract

The invention discloses a multi-sampling track inspection data processing model for an electric power inspection vehicle, and relates to the field of circuit inspection. The data processing model uses electric wires and their extension directions as the track inspections for automatic inspection by an electric power inspection vehicle in terms of data bottom-level logic. Based on images or videos provided by ground electric power inspection equipment and based on the collection of graphical data of the surrounding environment by a camera on the electric power inspection vehicle, a data algorithm is constructed in combination with mathematics and image processing techniques, and the status of electric wires is accurately analyzed and identified, so that the electric power inspection vehicle can perform high-precision automatic patrols based on the electric wires and their extension directions.

Description

Technical Field

[0001] The invention relates to the field of circuit inspection, and in particular to a multi-sampling track inspection data processing model for an electric power inspection vehicle. Background Art

[0002] Power line inspection is one of the important means to ensure the safe and reliable operation of the power system. In the traditional inspection method, inspectors need to climb up the poles or use ladders to conduct inspections, which poses safety hazards such as falling from heights and bad weather, and it is also impossible to conduct a comprehensive and detailed inspection of the power lines. In this technical context, in order to solve the problems existing in traditional power line inspections and improve inspection efficiency, artificial intelligence technology has been widely used in the field of power line inspections. With the rapid development of many cutting-edge artificial intelligence technologies such as holographic perception of equipment status, remote one-button sequential control of switching operations, and machine replacement of manual inspections, new ideas, algorithms, and technologies have gradually been applied to power line inspection operations, and have substantially improved the accuracy, efficiency, and intelligence level of identifying power line faults or defects.

[0003] Correspondingly, with the advancement and development of the above technologies, more and more power companies have begun to introduce intelligent inspection equipment and collect images of power lines and power equipment based on fixed cameras, panoramic acquisition and other image acquisition equipment. Among them, the drone inspection technology route is particularly prominent. It can be said that drone inspection technology has been widely used in power line inspection. Drone inspection can observe the ground situation at high altitude without blind spots, flexibly adjust the observation angle and position, and drones can also be remotely controlled to reduce the risk and cost of manual inspection. In harsh environments or high-risk tasks, drones can replace manual inspections, reducing safety risks.

[0004] But at the same time, drone inspections still have some shortcomings. For example, in some special environments, such as strong winds, rain, snow and other severe weather conditions, the flight stability and inspection results of drones will be seriously affected. At the same time, in the application practice of drone power line inspections, drones are often affected by factors such as electromagnetic interference during flight, especially considering that there are a large number of power equipment that present electromagnetic interference sources on the power line inspection path. This problem is often very serious and difficult to effectively overcome and eliminate. Another point that seriously restricts the application of drones in wide-area power line inspections is the existence of no-fly zones. For various reasons, wide-area and long-distance power line inspections often have no-fly restrictions in some areas, making it impossible for drones to inspect these areas.

[0005] Therefore, due to the above-mentioned defects of drone inspections, ground power inspection equipment (i.e., power line inspection vehicles, sometimes also referred to as power inspection robots in academia and applications) has gradually received attention and developed into one of the leading technical routes for intelligent inspection of power lines. At present, the leading technical route for ground power inspection equipment is that power inspection vehicles equipped with cameras use image acquisition to collect a large number of images or videos of power equipment, and then identify, eliminate or pre-check the risks of power line and power equipment failures based on built-in image recognition algorithms and fault recognition data models. The recognition results are finally reported to the data center and finally manually checked by power engineers or other operators.

[0006] Generally speaking, the power equipment image recognition algorithms loaded and pre-built in the above-mentioned ground power inspection equipment often have insufficient ability to extract image features and poor recognition accuracy, resulting in poor efficiency and accuracy in identifying power line and power equipment faults. This is often solved through later manual correction. With the advancement of technology, the efficiency and accuracy of equipment recognition are gradually improved, and the corresponding workload of manual later correction is getting smaller and smaller.

[0007] However, in addition to the above-mentioned issues about the efficiency and accuracy of recognition, there is another very important issue, namely: in order for the power line inspection car to be able to perform automatic inspections, it is first necessary to automatically identify and plan the power patrol trajectory. On this issue, the existing calculation model generally takes the wires and their extension direction as the basis, identifies the wires in the images collected by the power inspection car, and the direction of the wires, and then automatically constructs and plans the power patrol trajectory based on this and adjusts it in real time. Specifically, the existing calculation model generally treats the wires as standard straight lines, and the analytical expression of its graph after coordinateization has the characteristics of a linear function. In this way, for the images collected by the power inspection car, the calculation model can perform dynamic modeling and planning of the wires and the direction of the wires by identifying the primary analytical expression that spans the entire field of view (and continues to extend it beyond the field of view), and based on this, plan the patrol and perform automatic power line inspections.

[0008] However, there are some serious defects in the above calculation model. Simply put, for the images collected by the power inspection car, there are often many long straight graphic elements in its field of view, such as electric poles, tree trunks, and long straight stripes on the facades of buildings, which often lead to errors in the identification of real wires. In particular, considering the dynamic movement, bumps, and jitters of the inspection car, coupled with the complex interweaving of a large number of wires with trees, buildings and other backgrounds, wire identification errors and the resulting patrol planning errors often occur. The root cause of this error is that the primary analytical function recognition algorithm used in the existing calculation model is too simple. It only considers the geometric characteristics of the wire as a long straight line, without further algorithm planning and data modeling for detailed features such as the local curvature, concavity or symmetry of the wire.

[0009] Therefore, it is particularly important to develop an algorithm model that is compatible with the linear extension characteristics of power lines and incorporates their local curvature characteristics and curvature distribution structure in order to improve the automated patrol capabilities of ground power inspection equipment. Summary of the invention

[0010] The technical problem to be solved by the present invention is to provide a multi-sampling track inspection data processing model for an electric power inspection vehicle. On the basis of considering the linear extension characteristics of the power line, its local curvature characteristics and curvature distribution structure are incorporated into the constructed algorithm model to enhance the automated patrol capability of ground electric power inspection equipment.

[0011] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0012] A multi-sampling track inspection data processing model for an electric power inspection vehicle is characterized in that the data processing model uses the electric wires and their extension directions as the track inspections for the electric power inspection vehicle's automatic inspections in terms of data bottom-level logic, collects graphical data of the surrounding environment based on images or videos provided by ground electric power inspection equipment and cameras on the electric power inspection vehicle, constructs a data algorithm in combination with mathematics and image processing techniques, accurately analyzes and identifies the status of the electric wires, and enables the electric power inspection vehicle to perform high-precision automatic patrols based on the electric wires and their extension directions.

[0013] Furthermore, the data processing model identifies the wires and their extended directions in the surrounding environment pictures or videos collected by the power inspection vehicle, forming the data basis for the automatic inspection of the power inspection vehicle; its basic algorithm framework is: on the global data coordinates, the wires are identified as a single analytical straight line by the algorithm, thereby treating the graphic elements with a single analytical expression in the images or videos collected by the power inspection equipment as potential wires and potential automatic patrol trajectories; its core algorithm framework is: on the local refined data coordinates, the wires and towers are subjected to data intersection processing in the real physical space, and based on the data intersection of the wires and towers and their unique analytical curve characteristics, the interfering graphic elements in the potential wires and potential automatic patrol trajectories are eliminated, thereby realizing high-precision automatic patrol identification based on the wires and their extended directions.

[0014] Furthermore, in the core algorithm framework, the data intersection of the power lines and towers and their unique analytical curve characteristics include: A. local nonlinear curvature variation characteristics of the power lines or their analytical curves in the interval between two groups of power towers, and B. local nonlinear curvature variation characteristics of the power lines or their analytical curves near any power tower.

[0015] Furthermore, the local nonlinear curvature variation characteristics of the electric wire or its analytical curve in the interval between two groups of power towers may optionally include: the analytical curve of the electric wire forms a continuous, smooth, single concave, symmetrical curve between the two groups of power towers; wherein, A-α, the data characteristics corresponding to the continuity are: the analytical curve of the electric wire has a first-order finite derivative; A-β, the data characteristics corresponding to the smoothness are that the analytical curve of the electric wire has a second-order and above derivative; A-γ, the data characteristics corresponding to the single concave curve include: A-γ-ω has and only has one extreme point on the analytical curve of the electric wire between the two groups of power towers, that is, a point where the first-order derivative is zero, and This extreme point constitutes the local maximum point; A-γ-ψ has no inflection point on the analytical curve of the wire between the two groups of power towers, that is, the point where the second-order derivative is zero; A-γ-χ has its extreme point as the midpoint on the analytical curve of the wire between the two groups of power towers, and the line connecting any two points on both sides is higher than its extreme point; A-δ, the data feature corresponding to the symmetry is: the analytical curve of the wire has left-right equal symmetry according to a certain middle point; any one or several groups of local nonlinear curvature variation features are selected to perform high-precision identification of wires on the local refined coordinates; at the same time, it is compatible with the subsequent addition of other local curvature features to expand and improve the data model.

[0016] Furthermore, the local nonlinear curvature variation characteristics of the electric wire or its analytical curve near any power tower pole may optionally include: B-α, the analytical curve of the electric wire reaches a local minimum near any power tower pole; B-β, the non-smoothness of the analytical curve of the electric wire near any power tower pole; B-γ, the analytical curve of the electric wire has only a unilateral derivative near any power tower pole, that is, B-γ-ω, the analytical curve of the electric wire on the left side of the power tower pole has only a left derivative, and B-γ-ψ, the analytical curve of the electric wire on the side of the power tower pole has only a right derivative; any one or several groups of local nonlinear curvature variation characteristics are selected to perform high-precision identification of the electric wire on the local refined coordinates; and at the same time, it is compatible with the subsequent addition of other local curvature features to expand and improve the data model.

[0017] Furthermore, in terms of data processing, the line segment data of suspected electric wires sampled by the camera of the power inspection car is basically converted into a function curve, and its analytical geometry expression is recorded as f(x). The rate of change of the function curve, the curvature of the function curve and the convexity of the function are examined.

[0018] Further, optional data processing processes include:

[0019] a. Calculate the first-order differential df / dx of the function curve f(x) to obtain the slope characteristic function p(x);

[0020] b. Calculate the second-order differential d2f / dx2 of the function curve f(x) to obtain the data representation function g(x) of the curvature;

[0021] c. Analyze the extreme points of the function curve f(x) and the minimum points of the potential wire image curve;

[0022] d. Analyze the concavity and convexity of the potential wire image curve;

[0023] e. Analyze the symmetry of the potential wire image curve to determine whether it is symmetrical about a certain midpoint x0;

[0024] f. Rotation operation of the function curve f(x) according to the coordinate axis or other standard reference.

[0025] Furthermore, the extreme value point of the analysis function curve f(x) includes solving the equation df / dx=0, and determining the minimum value point of the wire image curve by comparing the extreme value points within the two groups of tower intervals, and determining the position of the minimum value point of the wire image curve as needed.

[0026] Furthermore, the analysis of the concavity and convexity of the potential wire image curve includes calculating the second-order derivative of the wire image curve f(x) within the range of two groups of tower poles, and confirming that the second-order derivative has sign consistency in the entire interval, that is, the second-order derivative is positive or negative in the entire interval, then the potential wire image curve exhibits single concavity / single convexity.

[0027] Furthermore, the confirming of the symmetry of the wire image curve and judging whether it is symmetric about a certain midpoint x0 includes selecting the midpoint x0 of the independent variable interval of the wire image curve f(x) and verifying whether the function values ​​of corresponding points in the left and right intervals are equal.

[0028] Compared with the prior art, the present invention has the following technical advances:

[0029] In general, the present invention effectively improves the automated patrol capability of ground power inspection equipment, and can automatically construct and plan power line inspection routes based on power lines and their extension directions, solving the problems of simple linear algorithm model recognition errors and poor adaptability to dynamic environments in the prior art, and providing more reliable and accurate technical support for intelligent inspection of power lines.

[0030] In the algorithm path, the basic algorithm framework of the present invention uses the wires as a first-order analytical straight line for algorithm recognition in the global data coordinates, and regards the graphic elements with a first-order analytical expression as potential wires and automatic patrol trajectories. Furthermore, the core algorithm framework eliminates interfering graphic elements based on the data intersection of wires and towers and their unique analytical curve characteristics in the local refined data coordinates, and realizes high-precision automatic patrol identification.

[0031] The core algorithm framework is mainly based on the local nonlinear curvature variation characteristics of power lines in data feature processing, including the continuous, smooth, single concavity, and symmetric curve characteristics of the power line between two groups of power towers, as well as the local maximum, non-smoothness, and unilateral derivative characteristics near any power tower. Based on the analysis of these local nonlinear characteristics, the corresponding data processing process (the analytic geometry expression corresponding to the graph after coordinateization) can be constructed: perform first-order and second-order differentials on the function curve f(x), analyze extreme points, concavity, and symmetry, etc. Through the analysis of these features (data features of the graphical analytical expression function), accurate dynamic identification of the power line can be ensured.

[0032] In addition, the algorithm model constructed by the present invention is compatible with the local curvature features added subsequently, and has good scalability and room for improvement.

[0033] The technical details and detailed technical effects of the present invention can be simultaneously referred to the specific implementation examples below. DETAILED DESCRIPTION

[0034] The following examples describe the present invention in detail. In the description of the following examples, specific details such as specific system structures, technologies, etc. are proposed for illustration rather than for limitation, so as to thoroughly understand the present application embodiments. However, it should be clear to those skilled in the art that the present application can also be implemented in other embodiments without these specific details. In other cases, the detailed description of well-known systems, devices, circuits and methods is omitted to prevent unnecessary details from hindering the description of the present application. It should be understood that when used in the present application specification and the attached claims, the term "comprising" indicates the presence of the described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or their collections. It should also be understood that the term "and / or" used in the present application specification and the attached claims refers to any combination of one or more of the items listed in association and all possible combinations, and includes these combinations.

[0035] In addition, in the description of the present application specification and the attached claims, the terms "first", "second", "third", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance. The reference to "one embodiment" or "some embodiments" described in the present application specification means that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in the different parts of this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0036] Embodiment 0, technical overview

[0037] Due to the many limitations of drone inspections in actual applications, such as no-fly zones and other issues that do not allow for roundabout solutions, and several other technical obstacles, ground-based power inspection equipment has gradually gained attention and developed into one of the leading technical routes for intelligent inspection of power lines.

[0038] Existing power inspection vehicles mainly rely on the recognition and planning of power lines and their extension directions to achieve automatic inspection (planning automatic inspection routes). Generally, the power lines are regarded as standard straight lines, and the patrol trajectory is constructed and planned by identifying the power lines and their directions in the image. However, this method has significant defects. Since the algorithm logic regards the power lines as standard straight lines, it is easy to misidentify the long straight stripes on the poles, tree trunks, and building facades as power lines. The inspection vehicle may bump and shake during dynamic movement, and the power lines are complexly intertwined with trees, buildings and other backgrounds, resulting in frequent wire recognition errors. The existing model only considers the geometric features of the wires and fails to incorporate detailed features such as the local curvature, convexity or symmetry of the wires. This is also the fundamental reason for the inaccurate patrol planning and the basis for improving the patrol algorithm in this study.

[0039] In view of this, the present invention not only considers the linear extension characteristics of power lines, but also incorporates its local curvature characteristics and curvature distribution structure into the constructed algorithm model to enhance the automatic patrol capability of ground power inspection equipment. The basic algorithm framework is: in the global data coordinates, the wires are regarded as a first-order analytical straight line for algorithm recognition, and the graphic elements with a first-order analytical expression are regarded as potential wires and automatic patrol trajectories; the core algorithm framework is: in the local refined data coordinates, based on the data intersection of wires and towers and their unique analytical curve characteristics, the interference graphic elements are eliminated to achieve high-precision automatic patrol recognition.

[0040] In the core algorithm framework, the data intersection of power lines and towers and their unique analytical curve features include: A. Local nonlinear curvature variation characteristics of power lines or their analytical curves in the interval between two groups of power towers, specifically including: the analytical curve of the power line forms a continuous, smooth, single concave, and symmetrical curve between the two groups of power towers; wherein, the data feature corresponding to continuity is: the analytical curve of the power line has a first-order finite derivative; the data feature corresponding to smoothness is that the analytical curve of the power line has a second-order and above derivative; the data features corresponding to the single concave curve include: ω has and only has one extreme point on the analytical curve of the power line between the two groups of power towers, that is, the point where the first-order derivative is zero, and this extreme point constitutes a local maximum point; ψ has no inflection point on the analytical curve of the power line between the two groups of power towers, that is, the point where the second-order derivative is zero; χ has no inflection point on the analytical curve of the power line between the two groups of power towers, that is, the point where the second-order derivative is zero; On the curve, with its extreme point as the midpoint, the line connecting any two points on both sides is higher than its extreme point; the data features corresponding to the symmetry are: the analytical curve of the wire has left-right equal-value symmetry according to a certain middle point; any one or several groups of local nonlinear curvature variation features are selected to perform high-precision identification of the wire on the local refined coordinates; and it is compatible with the subsequent addition of other local curvature features to expand and improve the data model; B. Local nonlinear curvature variation features of the wire or its analytical curve near any power tower, specifically including: the analytical curve of the wire reaches a local minimum near any power tower; the non-smoothness of the analytical curve of the wire near any power tower; the analytical curve of the wire has only a unilateral derivative near any power tower, that is, the analytical curve of the wire on the left side of the power tower has only a left derivative, and the analytical curve of the wire on the side of the power tower has only a right derivative.

[0041] In the algorithm construction of the model, one or several groups of local nonlinear curvature variation characteristics can be selected to perform high-precision identification of wires on the local refined coordinates; that is, based on the continuity, smoothness, single concavity, and symmetry of the wires between two groups of power towers, as well as the local maximum, non-smoothness, and unilateral derivative characteristics near any power tower, the corresponding data processing process is constructed, such as performing first-order differentiation and second-order differentiation on the function curve f(x), analyzing extreme points, concavity and convexity, and symmetry, etc. Through the analysis of these characteristics (data characteristics of the graphical analysis expression function), accurate dynamic identification of the wires can be ensured.

[0042] Of course, new curvature features and their distribution features can also be developed based on the actual physical state of the wires on the power lines, thereby subsequently adding other local curvature features to expand and improve the data model.

[0043] The following embodiments provide several examples of constructing algorithm models based on the local nonlinear curvature variation characteristics listed above.

[0044] Embodiment 1, original data processing

[0045] The line segments of suspected wires captured by the camera of the power inspection vehicle are embedded in the two-dimensional plane, and the two-dimensional coordinate system xoy is established. The wire image features are extracted and fitted into a wire image curve, denoted as f(x). The direction with the largest limit difference is taken as the independent variable x, and the wire image curve f(x) is obtained.

[0046] Example 2: Preparation of wire image curve model

[0047] On the basis of the above function curve, a new electric power inspection vehicle track inspection image recognition algorithm model was constructed through the joint development of multiple operating departments and data experts. The image recognition algorithm model first examines the rate of change of the above fitting function curve f(x) at each point, that is, the first-order differential df / dx of the function curve f(x) is calculated, thereby obtaining a data characterization function p(x) of how the slope of the initial function curve changes at the inspection point, that is, the slope characteristic of the wire image curve; the second step is to continue to examine how the slope of the desired function curve changes, that is, the degree of curvature of the measured function curve f(x), and the second-order differential of f(x) is calculated, thereby obtaining a data characterization function g(x) of the degree of curvature of the function curve f(x), that is, the degree of curvature of the wire image curve.

[0048] Embodiment 3: Processing of wire image curve features

[0049] Extreme point analysis: Take the first-order differential of the wire image curve f(x), solve df / dx=0, and there is only one zero solution x 0 That is, there is only one local maximum point of the function and only one minimum point on the overall wire curve. That is, the local minimum point of the function is the overall minimum point. In the natural state, the wire image curve has only one local maximum point and it is the minimum point. This section of wire has only one horizontal lowest point, which is a function feature of the wire image curve.

[0050] Extreme point analysis: The wire image curve is a continuous function and has no non-differentiable points, that is, the extreme point is the extreme point. The first-order differential of the wire image curve f(x) is obtained, that is, the slope function p(x) of the wire image curve. Let p = p(x), solve p = 0, and there is only one zero solution x = x 0 , and when x is greater than x 0 , p>0, that is, when the curve extends in the positive direction, the function value increases, the wire height increases, and x is less than x 0 , p<0, when the curve extends in the negative direction, the function value increases and the wire height increases. In addition, the wire graph curve is a continuous function and has no non-differentiable points. The above proves that the zero point x 0 is the minimum point, that is, the minimum point of the wire image curve. This section of wire has only one horizontal minimum point, which is a function feature of the wire image curve.

[0051] Example 4: Analysis of wire function properties

[0052] Function change rate analysis: Based on the natural state of the wire curve, the function model analysis is performed on the image feature of the wire image curve change rate. The first-order differential of the wire image curve f(x) is calculated, that is, the slope function p(x) of the wire image curve. Let p = p(x), x is greater than x 0 , p>0, that is, when the curve extends in the positive direction, the function value increases, the wire height increases, and x is less than x 0 , p<0, when the curve extends in the negative direction, the function value increases and the height of the wire increases. Specifically, the wire curve is strictly increasing when it extends to the left or right at the lowest point.

[0053] Slope change rate analysis: Based on the natural state of the wire curve, the curvature of the wire image curve is analyzed by a function model, and the second-order differential of the wire image curve f(x) is calculated, g(x) = d 2 f, let g = g(x), that is, the data representation function g(x) of the curvature of the wire image curve, x>x 0 Part,g>0,x <x 0 Part, g>0, extreme point x 0 The second-order differential values ​​on the left and right sides have the same sign and there is no inflection point, that is, when the wire curve is in a natural state, the line between any two points is above the arc segment of the wire image curve, which is specifically manifested in that the speed at which the wire image curve increases is getting faster and faster.

[0054] Function Concavity Analysis: Extract the concavity of the wire curve in its natural state, the image feature of the wire image curve, and use x 0 As the dividing point, the second-order differential of the wire image curve f(x) is calculated, g(x) = d 2 f, let g = g(x), the left and right sub-intervals are convex intervals, and the image characteristics of the wire curve include that the line between any two points is above the arc segment of the wire curve. And it satisfies that for any three points x on the interval 1 <x 2 <x 3 <x 0 , or x 0 >x 1 >x 2 >x 3 , satisfies the inequality f(x 2 )-f(x 1 ) / x 2 -x 1 ≤f(x 3 )-f(x 2 ) / x 3 -x 2. Specifically, the rate of change between any two points in the convexity interval of the wire image curve increases constantly.

[0055] Function symmetry analysis: If the pole structure type and horizontal height are the same, the function symmetry is good. The wire curve is in a natural state. The function model analysis is performed on the image feature of the wire image curve symmetry. The midpoint x is taken for the interval of the independent variable of the wire image curve. The function values ​​on the left and right intervals of x are completely equal. The wire image curve is about x=x. 1 Symmetric, that is, any range from x to x 1 The wire image curve has the same dependent variable (height) at the same distance, which is specifically manifested as the wire image curve about x=x 1 The axis is a complete mirror image of the two parts of the curve.

[0056] Embodiment 5: According to the image features of the wire image curve in a natural state, the wire image curve has only one extreme point.

[0057] In the sixth embodiment, according to the image characteristics of the wire image curve in the natural state, the extreme value point of the wire image curve is the minimum value point, and the minimum value point is also the minimum value point.

[0058] Example 7: If the pole structure type and horizontal height are the same, the wire image curve is in the natural state, and the midpoint x of the independent variable interval is 1 =x 0 .

[0059] From the above embodiments, it can be seen that at the application end, one or several groups of local nonlinear curvature variation features can be selected to perform high-precision identification of wires on local refined coordinates; that is, based on the continuity, smoothness, single concavity, and symmetry of the wires between two groups of power towers, as well as the local maximum, non-smoothness, and unilateral derivative characteristics near any power tower, a corresponding data processing process is constructed, such as performing first-order differentiation and second-order differentiation on the function curve f(x), analyzing extreme points, concavity, and symmetry, etc., and the analysis of these features (data features of the graphical analysis expression function) can ensure accurate dynamic identification of the wires. In terms of technical effects, the above-mentioned embodiment can more accurately identify the wires and their extension directions and reduce the misidentification rate by incorporating the local curvature characteristics and curvature distribution structure of the wires into the algorithm model; on the basis of the basic algorithm framework, the local nonlinear curvature variation characteristics of the wires between two groups of power towers and the local nonlinear curvature variation characteristics near any power tower are further introduced to make the wire identification more refined; based on the high-precision wire identification results, more accurate automatic track patrol planning can be achieved, thereby improving the automated patrol capability of ground power inspection equipment.

[0060] In the above embodiments, the description of each embodiment has its own emphasis. For the part that is not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. In each embodiment, the hardware implementation of the technology can directly adopt existing intelligent devices, including but not limited to industrial computers, PCs, smart phones, handheld stand-alone machines, floor-standing stand-alone machines, etc. Its input device preferably adopts a screen keyboard, its data storage and calculation module adopts existing memory, calculator, controller, its internal communication module adopts existing communication ports and protocols, and its remote communication adopts existing GPRS networks, World Wide Internet, etc.

[0061] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the above-mentioned method embodiment, which will not be repeated here. In the embodiments provided by the present invention, it should be understood that the disclosed device / terminal equipment and method can be implemented in other ways. For example, the device / terminal equipment embodiments described above are merely illustrative. For example, the division of modules or units is merely a logical function division. There may be other division methods in actual implementation, such as multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection of devices or units, which may be electrical, mechanical or other forms. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0062] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for processing multi-sampling track inspection data of an electric power inspection vehicle, characterized in that: The data processing method identifies the wires and their extension directions in the surrounding environment pictures or videos collected by the power inspection vehicle, forming the data basis for the automatic inspection of the power inspection vehicle; its basic algorithm framework is: in the global data coordinates, the wires are regarded as a first-order analytical straight line for algorithm recognition, thereby treating the graphic elements with a first-order analytical expression in the images or videos collected by the power inspection equipment as potential wires and potential automatic patrol trajectories; The core algorithm framework is: on the local refined data coordinates, the data of wires and towers are cross-processed in the real physical space, and based on the data cross-section of wires and towers and their unique analytical curve characteristics, the interference graphic elements in potential wires and potential automatic patrol trajectories are eliminated, and high-precision automatic patrol identification based on wires and their extended directions is achieved; In the core algorithm framework, the data intersection of the power lines and towers and their unique analytical curve features include: A. local nonlinear curvature variation characteristics of the power lines or their analytical curves in the interval between two groups of power towers, B. local nonlinear curvature variation characteristics of the power lines or their analytical curves near any power tower; The local nonlinear curvature variation characteristics of the electric wire or its analytical curve in the interval between two groups of power towers include: the analytical curve of the electric wire forms a continuous, smooth, single concave, symmetrical curve between the two groups of power towers; wherein, A-α, the data characteristics corresponding to the continuity are: the analytical curve of the electric wire has a first-order finite derivative; A-β, the data characteristics corresponding to the smoothness are that the analytical curve of the electric wire has a second-order and above derivative; A-γ, the data characteristics corresponding to the single concave curve include: A-γ-ω has and only has one extreme point on the analytical curve of the electric wire between the two groups of power towers, that is, The point where the first-order derivative is zero, and this extreme point constitutes the local maximum point; A-γ-ψ has no inflection point on the analytical curve of the wire between two groups of power towers, that is, the point where the second-order derivative is zero; A-γ-χ has its extreme point as the midpoint on the analytical curve of the wire between two groups of power towers, and the line connecting any two points on both sides is higher than its extreme point; A-δ, the data feature corresponding to the symmetry is: the analytical curve of the wire has left-right equal symmetry according to a certain middle point; select one or several groups of local nonlinear curvature variation features to perform high-precision identification of wires on local refined coordinates; The local nonlinear curvature variation characteristics of the electric wire or its analytical curve near any power tower include: B-α, the analytical curve of the electric wire reaches a local minimum near any power tower; B-β, the non-smoothness of the analytical curve of the electric wire near any power tower; B-γ, the analytical curve of the electric wire has only a unilateral derivative near any power tower, that is, B-γ-ω, the analytical curve of the electric wire on the left side of the power tower has only a left derivative, and B-γ-ψ, the analytical curve of the electric wire on the side of the power tower has only a right derivative; any one or several groups of local nonlinear curvature variation characteristics are selected to perform high-precision identification of the electric wire on the local refined coordinates.

2. According to claim 1, a method for processing multi-sampling track inspection data of an electric power inspection vehicle is characterized in that: In data processing, first, the line segment data of the suspected electric wires sampled by the camera of the power inspection vehicle is converted into a function curve, and its analytical geometry expression is denoted as f(x).

3. According to claim 2, a method for processing multi-sampling track inspection data of an electric power inspection vehicle is characterized in that: In terms of data processing, further optional data processing processes include: a. Calculate the first-order differential df / dx of the function curve f(x) to obtain the slope characteristic function p(x); b. Find the second-order differential d of the function curve f(x) 2 f / dx 2 , get the data representation function g(x) of the bending degree; c. Analyze the extreme points of the function curve f(x) and the minimum points of the potential wire image curve; d. Analyze the concavity and convexity of the potential wire image curve; e. Analyze the symmetry of the potential wire image curve to determine whether it is symmetrical about a certain midpoint x0.

4. The method for processing multi-sampling track inspection data of a power inspection vehicle according to claim 3 is characterized in that: The extreme value point of the analytical function curve f(x) includes solving the equation df / dx=0, and determining the minimum value point of the wire image curve by comparing the extreme value points within the two groups of tower intervals, and determining the position of the minimum value point of the wire image curve as needed.

5. According to claim 3, a method for processing multi-sampling track inspection data of an electric power inspection vehicle is characterized in that: The analysis of the concavity and convexity of the potential wire image curve includes obtaining the second-order derivative of the wire image curve f(x) within the range of two groups of tower poles, and confirming that the second-order derivative has the sign consistency of the entire interval, that is, the second-order derivative is positive or negative in the entire interval, then the potential wire image curve exhibits single concavity / single convexity.

6. The method for processing multi-sampling track inspection data of a power inspection vehicle according to claim 3 is characterized in that: The confirming the symmetry of the wire image curve and judging whether it is symmetric about a certain midpoint x0 includes selecting the midpoint x0 of the independent variable interval of the wire image curve f(x) and verifying whether the function values ​​of corresponding points in the left and right intervals are equal.