Zero-value insulator detection method and system based on voltage distribution characteristics

Abstract

The invention discloses a zero-value insulator detection method and system based on voltage distribution characteristics, and the method comprises the steps: building a three-dimensional simulation model according to the structure parameters of an inspection unmanned plane, the material parameters of an insulator chain and an arrangement mode, and calculating equivalent distributed capacitance through finite element numerical simulation; building an insulator detection equivalent circuit model through the calculated capacitance parameters, and changing an insulation resistance value to calculate a voltage distribution value of an insulator to obtain a training sample; then, establishing a random forest model, and substituting training samples such as voltage distribution values and corresponding insulation resistance values into the model to complete training of the model; and finally, inputting the voltage distribution characteristics of the actually-measured insulator, outputting the resistance value of the detected insulator through model diagnosis, and judging whether the detected insulator is a zero-value insulator or not according to the resistance value. According to the method, the influence of an unmanned aerial vehicle measurement probe and a measurement loop on insulator voltage distribution is considered, the interval of the actual insulator resistance value is obtained through the random forest model, and the detection accuracy is improved.

Description

Technical Field

[0001] The invention belongs to the field of zero-value insulator detection of transmission lines, and in particular relates to a zero-value insulator detection method and system based on voltage distribution characteristics. Background Art

[0002] Insulators are exposed to complex and harsh environments for a long time and are easily corroded by pollutants such as wind, sun, acid rain, thunderstorms, etc. As the operating time increases, the insulation performance of individual insulators will continue to decrease, and the insulation resistance will approach 0, affecting the safe operation of transmission lines and posing a hidden danger to people's lives. Therefore, it is necessary to regularly detect deteriorated porcelain insulators.

[0003] Traditional detection methods mainly use insulation resistance method, voltage distribution method, spark gap method, etc. to identify insulators. They are mainly manual detection. Such methods are cumbersome and require tower climbing, which has a high risk factor. The insulator inspection drone flies along the insulator string and collects the voltage value at both ends of the insulator during the flight, which can more conveniently obtain the voltage value at both ends of the insulator. However, there are also the following two problems in measuring potential using inspection drones: on the one hand, the inspection drone will affect the measurement result of the voltage value at both ends of the insulator due to the influence of its own measurement resistance and the corresponding probe structure during the voltage measurement process; on the other hand, the traditional method has difficulty in determining the threshold value for the comparison of potential distribution, and there is a problem of misjudgment. Therefore, it is necessary to propose a faster and more accurate method to calculate the insulation resistance of different insulators based on the collected voltage distribution characteristics to detect the zero-value insulators of the line.

[0004] Prior art document 1 (CN107589351 A, application publication date 2018.01.16) discloses a low- and zero-value insulator detection method for an insulator detection robot. A three-dimensional model is established according to the position of the insulator string, the structural dimensions and material parameters of the insulator string and the insulator detection robot, and the distributed capacitance matrix is ​​calculated in the electrostatic field using the finite element method; an overall circuit model for insulator detection is built according to the distributed capacitance parameters and the measurement loop parameters; low- and zero-value insulators are simulated by changing different insulation resistance values, and the voltage distribution characteristics of the insulator string under different conditions are calculated; the measurement results are compared with the simulation results to determine whether low- and zero-value insulators are present and their positions.

[0005] Compared with the prior art document 1, the present invention takes into account the influence of the inspection drone measurement probe loop on the voltage distribution characteristics of the insulator string, the simulation calculation results are more accurate, and the performance of the random forest model established based on this is more superior; however, the prior art document 1 does not take this into account.

[0006] The present invention is based on the voltage distribution characteristics and combined with the random forest algorithm model to diagnose the resistance interval of the insulator under test, and quantitatively analyzes whether it is a zero-value insulator through the threshold value, so the diagnosis result is more accurate. However, technical document 1 simply compares the measurement results with the simulation results, which has a large error. Summary of the invention

[0007] In order to solve the deficiencies in the prior art, the present invention provides a zero-value insulator detection method and system based on voltage distribution characteristics, which solves the technical problem in the prior art that the inspection drone will affect the measurement result of the voltage value at both ends of the insulator due to the influence of its own measurement resistance and the corresponding probe structure during the voltage measurement process.

[0008] The present invention adopts the following technical solution.

[0009] The present invention provides a zero-value insulator detection method based on voltage distribution characteristics, which is characterized by comprising:

[0010] A three-dimensional model of the transmission line including an insulator inspection drone was established, and the equivalent distributed capacitance was simulated and calculated through finite element analysis. An equivalent circuit model for insulator detection was built, the voltage distribution values ​​of insulators with different degrees of degradation were calculated, and the original sample set was constructed. The random forest model was established using the Bagging algorithm and the model was trained. After the model training was completed, the insulator to be tested was input, and the random forest model determined whether the insulator to be tested was a zero-value insulator based on the resistance value.

[0011] Preferably, a three-dimensional model of the transmission line is built according to the structural parameters of the inspection drone, tower, conductor, crossarm and insulator, the potential is increased at the uppermost connecting fitting of the insulator string, and the potential of the steel foot and the rigid cap are coupled.

[0012] Preferably, the boundary conditions of the three-dimensional model of the power transmission line include electric field boundary conditions, boundary conditions applied to the outer air domain boundary, and high-voltage electrode and insulator boundary conditions.

[0013] Electric field boundary conditions:

[0014]

[0015] in, represents the electric field gradient, is the electric field point, x, y, z represent the three-dimensional coordinates in the three-dimensional electric field model;

[0016] Apply boundary conditions to the outer air domain boundary:

[0017]

[0018] Boundary conditions between high voltage electrodes and insulators:

[0019] is the input high potential value, U1 is the high voltage electrode voltage.

[0020] Preferably, in the process of establishing the three-dimensional model of the transmission line, an inner air domain is first established to wrap the insulator, and an outer air domain is set outside the inner air domain to wrap the entire three-dimensional model of the transmission line. A simulation calculation model is established according to the boundary conditions of the three-dimensional model of the transmission line, and the insulator string capacitance parameters and the equivalent capacitance value of the electric field probe of the inspection drone are obtained through iterative solution.

[0021] Preferably, the voltage level corresponds to the number of insulators. When the number of insulators is 1, the voltage is 6-10kV; when the number of insulators is 3, the voltage is 35kV; when the number of insulators is 7, the voltage is 110kV; when the number of insulators is 11, the voltage is 220kV.

[0022] Preferably, when the insulation resistance value of the insulator is lower than 10 megohms, it is judged as a zero-value insulator; when the insulation resistance value of the insulator is 10-300 megohms, it is judged as a low-value insulator; when the voltage level is below 500kV, the insulation resistance of the insulator is not less than 300 megohms, it is judged as a degraded insulator; when the voltage level is 500kV, the insulation resistance of the insulator is not less than 500 megohms, it is judged as a degraded insulator.

[0023] Preferably, the normalized insulator voltage distribution value, ambient temperature, humidity and voltage level are used as the original sample set of the random forest model, and multiple training samples are randomly selected with replacement from the original sample set. The sample size each time is 3 / 4 of the original sample, and the sampling is randomly replaced. A regression decision subtree is constructed for each training sample, and the output of the decision subtree is the resistance range of each insulator. The convergence error of the parameter calculation model is changed to an error threshold of 0.1%.

[0024] The present invention also provides a zero-value insulator detection system based on voltage distribution characteristics, the system is a system used by the above-mentioned zero-value insulator detection method based on voltage distribution characteristics, comprising:

[0025] Transmission line module, including inspection drone unit, tower unit, conductor unit, cross arm unit and insulator unit;

[0026] A drone inspection module, comprising a drone probe unit and an equivalent capacitance unit, wherein the drone probe unit is used to detect the voltage of an insulator string;

[0027] The insulator low zero value detection module includes a decision tree unit and a model diagnosis unit, which is used to diagnose whether the insulator string is a low zero value.

[0028] The present invention also provides a terminal, including a processor and a storage medium; the characteristics are:

[0029] The storage medium is used to store instructions;

[0030] The processor is configured to operate according to the instructions to execute the steps according to the aforementioned method.

[0031] The present invention also provides a computer-readable storage medium on which a computer program is stored, characterized in that the program implements the steps of the above method when executed by a processor.

[0032] The beneficial effect of the present invention is that, compared with the prior art,

[0033] (1) This method takes into account the influence of the inspection drone measurement probe loop on the voltage distribution characteristics of the insulator string. The simulation results are more accurate, and the performance of the random forest model established based on this method is more superior.

[0034] (2) Based on the calculation using the finite element method, this method accurately obtains the distribution parameters of each component of the insulator string, pole tower and insulator inspection drone. The resistance values ​​at different positions of the same insulator string are different. At the same time, the inspection drone probe measures the first to nth insulators along the insulator, and the position of the drone is also different. In each case, a corresponding insulator inspection drone, pole tower and insulator three-dimensional model is built according to the corresponding position. The resistance values ​​at different positions are obtained according to the finite element numerical simulation method in step 1, and then the potential distribution of the insulator string at different positions under different low and zero value insulators is accurately simulated and calculated. This method not only improves the calculation efficiency, but also builds a more comprehensive and detailed database for the detection of low and zero value insulators.

[0035] (3) This method is based on the voltage distribution characteristics and combines the random forest algorithm model to diagnose the resistance range of the insulator being tested. Through the threshold quantitative analysis, whether it is a zero-value insulator is determined, and the diagnosis result is more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 The following is a flow chart of a method for detecting zero-value insulators based on voltage distribution characteristics of the present invention:

[0037] Figure 2 A schematic diagram of a three-dimensional physical model of an insulator string according to an embodiment of the present invention;

[0038] Figure 3 A schematic diagram of an equivalent circuit model for insulator detection according to an embodiment of the present invention;

[0039] Figure 4 A schematic diagram of the distributed voltage of an insulator string obtained by simulating low and zero value insulators according to an embodiment of the present invention. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, other embodiments obtained by ordinary technicians in this field without creative work are all within the scope of protection of the present invention.

[0041] like Figure 1 As shown, embodiment 1 of the present invention provides a zero-value insulator detection method based on voltage distribution characteristics, comprising:

[0042] Step 1: Establish a three-dimensional model of the transmission line including the insulator inspection drone, and simulate and calculate the equivalent distributed capacitance through finite element analysis;

[0043] A three-dimensional model of the transmission line is built based on the structural parameters of the inspection drone, tower, conductor, crossarm, and insulator. After the model is built, the parameters of each part are set, the potential is increased at the top connection fitting of the insulator string, and the potential of the steel foot and the steel cap are coupled. When calculating the electric field around the insulator, the electric field calculation belongs to the category of electrostatic field, which is calculated by Introducing the scalar potential function, in the three-dimensional electric field model, the boundary conditions of the system include the electric field boundary conditions, the boundary conditions imposed on the outer air domain boundary, and the high-voltage electrode and insulator boundary conditions. The electric field boundary conditions are:

[0044]

[0045] in, represents the electric field gradient, is the electric field point, x, y, z represent the three-dimensional coordinates in the three-dimensional electric field model, and the boundary conditions are applied to the outer air domain boundary:

[0046]

[0047] Boundary conditions between high voltage electrodes and insulators:

[0048] is the input high potential value, U1 is the high voltage electrode voltage;

[0049] The insulator geometry is obtained by measuring its true shape, and then the mesh is divided. Due to the characteristics of the insulator height, high-voltage electrode, calculation area and insulator shed, and in order to meet the calculation accuracy and convergence requirements, a multi-level meshing method is adopted here to mesh the calculation model. The model is divided into a series of small units.

[0050] During the modeling process, considering the electric field distribution around the insulator, it is necessary to establish an inner air domain to wrap the insulator, and an outer air domain needs to be added outside the inner air domain to wrap the entire model. After the calculation model is meshed, the maximum size of the outer air domain mesh reaches 15mm, and the minimum size is 5mm; while the maximum size of the insulator part is 4mm, and the minimum size is only 1mm. Then the mathematical model and equations are established, the model is established through the electromagnetic equation, and the boundary conditions are applied, including the upper and lower limits of the voltage. Finally, the equivalent capacitance value is obtained by iterative solution. The numerical method adopts the least squares method, with the minimization target as the goal, and is solved by Newton nonlinear iteration. Finally, the mathematical equations are solved by numerical methods, and the capacitance parameters of the insulator string and the equivalent capacitance value of the electric field probe of the inspection drone are solved by iterative methods.

[0051] Depend on Introduce the scalar potential function, that is

[0052]

[0053]

[0054] Among them, E represents the electric field intensity, ψ represents the magnitude of magnetic flux, D represents the electric displacement vector, and р represents the equivalent capacitance parameter of the insulator string;

[0055]

[0056] Among them, B represents the magnetic induction intensity, l represents the Gaussian closed curve, S represents the Gaussian closed surface, and p V represents the charge density per unit volume, and V represents the charge per unit volume;

[0057] From D = εE, we can get the objective function of the electrostatic field control equation:

[0058]

[0059] Finally, the equivalent capacitance value is obtained by iterative solution, where ε0 is the vacuum dielectric constant, ε r is the relative dielectric constant, ε is the dielectric constant, and φ is the magnetic flux.

[0060] The numerical method uses the least square method with the goal of minimizing the problem and solves it through Newton nonlinear iteration.

[0061] Step 2: Build an equivalent circuit model for insulator detection, and calculate the voltage distribution value of the insulator by changing different insulation resistance values ​​in the model;

[0062] The implementation process includes: establishing an insulator string parameter model based on the insulator string capacitance parameters calculated in step 1 and the equivalent capacitance value of the electric field probe of the inspection drone; the insulator string parameter model is as follows: Figure 4As shown in the figure, in order to simulate insulators with different degrees of deterioration, the resistance values ​​in the model are changed to 10MΩ, 50MΩ, 100MΩ, 150MΩ, 200MΩ, and 250MΩ. A 220kV voltage is applied to both ends of the model, and the voltage values ​​at both ends of each insulator when the inspection drone inspects along the insulator string are simulated to obtain the distributed voltage of the entire string of insulators.

[0063] Step 3: Based on the insulator voltage distribution value, ambient temperature, humidity and voltage level as feature inputs and the insulator resistance range as model output, a random forest model is established using the Bagging algorithm and trained with the simulated samples;

[0064] The voltage level corresponds to the number of insulators. When the number of insulators is 1, the voltage is 6-10kV; when the number of insulators is 3, the voltage is 35kV; when the number of insulators is 7, the voltage is 110kV; when the number of insulators is 11, the voltage is 220kV.

[0065] ① Voltage level 500kV, the insulation resistance of the insulator is not less than 500MΩ, which is judged as a deteriorated insulator;

[0066] ② For voltage levels below 500kV, if the insulation resistance of the insulator is not less than 300MΩ, it is judged as a degraded insulator.

[0067] Generally, deteriorated insulators with an insulation resistance of 10MΩ or more are called low-value insulators, while those with an insulation resistance of less than 10MΩ are called zero-value insulators.

[0068] Therefore, it is specifically divided into four ranges: 0-10 megohms, 10-300 megohms, 300-500 megohms, and 500 megohms and above to distinguish low zero value insulators, which represent zero value insulators, low value insulators, and degraded insulators respectively.

[0069] The implementation process includes: using the Bagging algorithm to take the normalized insulator voltage distribution value, ambient temperature, humidity and voltage level as feature input, randomly extracting several training samples from the original samples with replacement, the sample size extracted each time is 3 / 4 of the original sample, and the sampling is randomly replaced, and a regression decision subtree is constructed for each training sample. The output of the decision subtree is the resistance range of each insulator. By changing different parameter values ​​and obtaining the corresponding model convergence error, this process is repeated until the error convergence training is completed.

[0070] Step 4: Finally, by inputting the voltage distribution characteristics of the measured insulator, the resistance value of the measured insulator is output through model diagnosis, and it is determined whether the measured insulator is a zero-value insulator based on the resistance value;

[0071] The implementation process includes: after inputting the voltage distribution characteristics and related parameters of the insulator, the random forest model outputs the insulator resistance, which is compared with the critical value of the zero-value insulator resistance. When the measured insulator resistance is greater than or equal to the critical value of the zero-value insulator resistance, it is a normal insulator; when the measured insulator resistance is less than the critical value of the zero-value insulator resistance, it is a zero-value insulator.

[0072] Embodiment 2 of the present invention provides a zero-value insulator detection system based on voltage distribution characteristics, the system is a system used by the aforementioned zero-value insulator detection method based on voltage distribution characteristics, and includes:

[0073] Transmission line module, including inspection drone unit, tower unit, conductor unit, cross arm unit and insulator unit;

[0074] A drone inspection module, comprising a drone probe unit and an equivalent capacitance unit, wherein the drone probe unit is used to detect the voltage of an insulator string;

[0075] The insulator low zero value detection module includes a decision tree unit and a model diagnosis unit, which is used to diagnose whether the insulator string is a low zero value.

[0076] Example 3

[0077] This example uses a 220kV insulator string detection test conducted in a laboratory as an example to illustrate the implementation method of this patent:

[0078] 1) The main components of the insulator inspection drone include the drone body, pod, space electric field probe, ranging module, detection mainboard, handheld terminal and lithium battery and other modules; the degraded insulator detection system structure includes three parts: drone body, ground terminal and airborne terminal. Among them, the drone body plays the role of a carrying platform. The device on the airborne terminal completes the detection. The ground terminal processes and displays the data collected by the airborne terminal. The ground terminal and the airborne terminal transmit data through a wireless communication device. The drone body consists of a detachable gimbal, arm, brushless motor, propeller, fuselage, GPRS module, RTK module and forward-looking system. The forward-looking system is installed at the front of the fuselage, including the forward-looking infrared perception system and the forward-looking vision system. The RTK module and GPRS module are installed on the top of the fuselage. The control mainboard is installed inside the fuselage. The gimbal is installed at the bottom of the fuselage. The motor and propeller are installed on the arm. The propeller is fixed on the output shaft of the motor. The motor is connected to the control mainboard through an electronic speed regulator. The inspection drone model is as follows: Figure 2 shown.

[0079] The insulator model selects the xwp-60 double-umbrella porcelain insulator string in the 220kV voltage level line for modeling. The xwp-60 structure height is 165mm, the creepage distance is 300mm, the nominal disc diameter is 255mm, and the rigid cap diameter is 128mm. In the laboratory, a steel frame is used to simulate the tower crossarm. The steel frame size is 2200mm×1200mm×120mm. The established insulator string model consists of 15 xwp-70 insulators. The measurement points are at the same distance from the center of each insulator. They are numbered 1-15 starting from the high-voltage end to explore the spatial electric field characteristics piece by piece. Based on the xwp-70 double-umbrella suspension porcelain insulator string under the 220kV voltage level, the influence of factors such as towers, hardware, conductors and lightning conductors are also considered in the modeling process, and the modeling is carried out with the help of SolidWorks software. When collecting the insulator electric field, the simulation model of the insulator string is as follows Figure 3 shown.

[0080] 2) Under the framework of the constructed three-dimensional model, accurate parameter configuration was performed according to the material properties of each component, and the insulators were marked from 1 to 15. In order to ensure the authenticity of the simulation environment, the ambient atmospheric pressure was set to 101kPa, the temperature was constant at 25℃, the humidity was controlled at 51%, and the surface of the insulator was ensured to be dry. In terms of material parameters, the relative dielectric constant of the air medium was set to 1, the relative dielectric constant of porcelain was set to 5.41, and the relative dielectric constant of cement was set to 14. The entire model was carefully meshed, with a total of 1.3 million mesh units. In the setting of the model boundary conditions, the rated voltage of 220kV was loaded at key locations such as the conductor and the lower hardware, and the ground potential was loaded at the corresponding locations such as the tower and the upper hardware. Since the normal working voltage of the transmission line insulator is 50Hz power frequency alternating current, it can be considered that the voltage at both ends of the insulator is constant in a short time. Therefore, when studying the electric field distribution of the insulator, the insulator electric field is treated as an electrostatic field. The electrostatic field simulation was performed using the advanced finite element method to obtain the capacitance distribution parameters of the insulator string. When calculating the electric field around the insulator, the electric field calculation belongs to the category of electrostatic field and can be analyzed using the following model.

[0081]

[0082] The electrostatic field control equation is obtained:

[0083]

[0084] In the 3D electric field model, there are electric field boundary conditions:

[0085]

[0086] Apply boundary conditions to the outer air domain boundary:

[0087]

[0088] Boundary conditions between high voltage electrode and load insulator:

[0089] is the input high potential value, U1 is the high voltage electrode voltage.

[0090] Due to the characteristics of insulator height, high-voltage electrode, calculation area, and insulator shed, and in order to meet the calculation accuracy and convergence requirements, a multi-level grid division method is adopted to grid the calculation model.

[0091] During the modeling process, considering the electric field distribution around the insulator, an inner air domain needs to be established to wrap the insulator, and an outer air domain needs to be added outside the inner air domain to wrap the entire model. After the calculation model is meshed, the maximum size of the outer air domain mesh reaches 15mm, and the minimum size is 5mm; while for the insulator part, the maximum size is 4mm, and the minimum size is only 1mm.

[0092] Finally, an insulator string detection circuit model was built, including three-dimensional modeling, grid division, and the least squares method was used to iteratively solve the above two parameters in combination with boundary conditions, as shown in Table 1, in pF. The insulator monitoring equivalent circuit is as follows: Figure 4 shown.

[0093]

[0094]

[0095] 3) In order to obtain the voltage distribution characteristics of low and zero value insulators, the line model data obtained by the insulator inspection drone at different positions is used to simulate the voltage values ​​at both ends of each insulator when the drone flies along the insulator string to obtain the voltage distribution of the entire string of insulators. The fifth insulator is selected to simulate the low and zero value insulator, and the insulation resistance of the fifth insulator is set to 10MΩ, 50MΩ, 100MΩ, 150MΩ, 200MΩ, 250MΩ, and 300MΩ, respectively, and the insulation distribution voltage is calculated respectively. When the insulation resistance drops below 150MΩ, that is, it enters the low value range, obvious voltage changes can be observed through the voltage distribution characteristics.

[0096] 4) Carry out simulation calculations for various low and zero value insulators, and use the distributed voltage calculation results to establish a database. A decision tree model is established based on the insulator voltage distribution value, ambient temperature, humidity and voltage level as feature inputs and the insulator resistance range as output.

[0097] Set the T normalized samples as the original samples, randomly draw multiple training samples from the original samples with replacement. The number of samples drawn each time is 3 / 4 of the original samples, and the random sampling with replacement is performed. Then, construct a regression decision tree for each training sample respectively, and finally form a decision tree. After determining the sample category n and the regression decision tree, calculate the regression value of each decision subtree, where θ t is a random variable that follows an independent and identical distribution. Finally, take the mean of the regression values of multiple models as the regression prediction value to generate a random forest model. The formula is as follows:

[0098]

[0099] For the regression prediction method, the voting method is adopted in this paper. Assume that the ensemble contains M base learners (h1, h2, …, h M ), and the set of categories is {k1, k2, …, k n}. Denote the output of the i-th base learner h i on the sample n as an N-dimensional vector (h i 1 (n), h i 2 (n), …, h i N (n)), where h i j (n) represents the output of the classifier on the category a j . If a certain interval gets more than half of the votes, it is predicted as that category; otherwise, it is rejected. The operation formula is:

[0100]

[0101] Loop this process and calculate its error to check if it meets the requirements. The specific formula for the error Error is:

[0102]

[0103] where N is the number of samples in the training data.

[0104] By changing different values of n and θ t and obtaining the corresponding model convergence error, when Error last - Error new < a, take the current values of n and θ t as the final parameters. Here, Error last represents the previous error value, Error new represents the current error value, and a represents the error threshold, which is taken as 0.1% here

[0105] The number of decision trees is determined according to the value of the regression decision subtree, the normalized training samples are substituted into the random forest model, and the decision tree diagram is drawn to complete the establishment and training of the random forest model.

[0106] 5) The voltage distribution characteristics and environmental parameters collected by the inspection drone are sent to the random forest model, and the model diagnoses the range of resistance values ​​of each insulator. According to the requirements of DL / T596-1996 "Preventive Test Procedures for Power Equipment", the insulation resistance of 220kV suspension insulators should not be less than 150 megohms. Those below the above level are judged as low-quality insulators, and the location of low-value insulators is finally determined.

[0107] Specifically, after inputting the voltage distribution characteristics, ambient temperature, humidity and voltage level of the insulator, the random forest model outputs the resistance range of the insulator, which is compared with the critical value of the zero-value insulator resistance. When the resistance of the insulator under test is greater than or equal to the critical value of the zero-value insulator resistance, it is a normal insulator. When the resistance of the insulator under test is less than the critical value of the zero-value insulator resistance, it is a zero-value insulator. The detection of the insulator under test is now completed.

[0108] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which all fall within the protection of the present invention.

[0109] Example 4

[0110] In order to study the influence of the drone body and electric field probe on the electric field distribution around the insulator string when the drone flies in, a three-dimensional simulation model of the drone is established. The relative dielectric constants of the drone body and the insulation part of the electric field probe and the electrode on the probe surface are 6 and 500 respectively. In the simulation analysis, when the drone flies in with the electric field probe, the electric field distribution characteristics around the insulator string are analyzed at a distance of 10cm, 20cm, 30cm, 40cm, 50cm and 60cm from the center axis of the insulator string. The simulation results show that when the distance to the insulator string is 10cm, the voltage distribution characteristics are quite different, with a gap of 2.3kV / m. When the horizontal offset distance is 40cm, the gap is reduced to 0.17kV / m. When the horizontal offset distance is 60cm, the gap is reduced to 0.36kV / m. In the above six cases, the accuracy rates of the four simulations are 72.3%, 78.5%, 88.6%, 93.1%, 91.1% and 87.4% respectively. Compared with the existing technical documents, the accuracy rate is greatly improved.

[0111] The beneficial effect of the present invention is that, compared with the prior art,

[0112] (1) This method takes into account the influence of the inspection drone measurement probe loop on the voltage distribution characteristics of the insulator string. The simulation results are more accurate, and the performance of the random forest model established based on this method is more superior.

[0113] (2) Based on the calculation using the finite element method, this method accurately obtains the distribution parameters of each component of the insulator string, pole tower and insulator inspection drone. The resistance values ​​at different positions of the same insulator string are different. At the same time, the inspection drone probe measures the first to nth insulators along the insulator, and the position of the drone is also different. In each case, a corresponding insulator inspection drone, pole tower and insulator three-dimensional model is built according to the corresponding position. The resistance values ​​at different positions are obtained according to the finite element numerical simulation method in step 1, and then the potential distribution of the insulator string at different positions under different low and zero value insulators is accurately simulated and calculated. This method not only improves the calculation efficiency, but also builds a more comprehensive and detailed database for the detection of low and zero value insulators.

[0114] (3) This method is based on the voltage distribution characteristics and combines the random forest algorithm model to diagnose the resistance range of the insulator being tested. Through the threshold quantitative analysis, whether it is a zero-value insulator is determined, and the diagnosis result is more accurate.

[0115] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0116] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples of computer-readable storage media (a non-exhaustive list) include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the foregoing. As used herein, a computer-readable storage medium is not to be interpreted as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.

[0117] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0118] The computer program instructions for performing the operation of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. Computer-readable program instructions may be executed completely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be customized by utilizing the state information of the computer-readable program instructions, and the electronic circuit may execute the computer-readable program instructions, thereby realizing various aspects of the present disclosure.

[0119] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A zero-value insulator detection method based on voltage distribution characteristics, characterized in that: include: A three-dimensional model of the transmission line including an insulator inspection drone was established, and the equivalent distributed capacitance was simulated and calculated using finite element analysis; An insulator detection equivalent circuit model was built, the voltage distribution values ​​of insulators with different degrees of degradation were calculated, the original sample set was constructed, and the random forest model was established using the Bagging algorithm to train the model. After the model training was completed, the insulator to be tested was input, and the random forest model determined whether the insulator to be tested was a zero-value insulator based on the resistance value.

2. A method for detecting zero-value insulators based on voltage distribution characteristics according to claim 1, characterized in that: A three-dimensional model of the transmission line is built based on the structural parameters of the inspection drone, tower, conductor, crossarm and insulator. The potential is increased at the top connecting fitting of the insulator string, and the potential of the steel foot and the rigid cap are coupled.

3. A method for detecting zero-value insulators based on voltage distribution characteristics according to claim 2, characterized in that: The boundary conditions of the three-dimensional model of the transmission line include electric field boundary conditions, boundary conditions imposed on the outer air domain boundary, and high-voltage electrode and insulator boundary conditions. Electric field boundary conditions: in, represents the electric field gradient, is the electric field point, x, y, z represent the three-dimensional coordinates in the three-dimensional electric field model; Apply boundary conditions to the outer air domain boundary: Boundary conditions between high voltage electrodes and insulators: is the input high potential value, U1 is the high voltage electrode voltage.

4. A method for detecting zero-value insulators based on voltage distribution characteristics according to claim 3, characterized in that: In the process of establishing the three-dimensional model of the transmission line, an inner air domain is first established to wrap the insulator, and an outer air domain is set outside the inner air domain to wrap the entire three-dimensional model of the transmission line. A simulation calculation model is established according to the boundary conditions of the three-dimensional model of the transmission line, and the insulator string capacitance parameters and the equivalent capacitance value of the electric field probe of the inspection drone are obtained through iterative solution.

5. A method for detecting zero-value insulators based on voltage distribution characteristics according to claim 4, characterized in that: The voltage level corresponds to the number of insulators. When the number of insulators is 1, the voltage is 6-10kV; when the number of insulators is 3, the voltage is 35kV; when the number of insulators is 7, the voltage is 110kV; when the number of insulators is 11, the voltage is 220kV.

6. A method for detecting zero-value insulators based on voltage distribution characteristics according to claim 5, characterized in that: When the insulation resistance value of the insulator is lower than 10 megohms, it is judged as a zero-value insulator; when the insulation resistance value of the insulator is 10-300 megohms, it is judged as a low-value insulator; when the voltage level is below 500kV, the insulation resistance of the insulator is not less than 300 megohms, it is judged as a degraded insulator; when the voltage level is 500kV, the insulation resistance of the insulator is not less than 500 megohms, it is judged as a degraded insulator.

7. A method for detecting zero-value insulators based on voltage distribution characteristics according to claim 6, characterized in that: The normalized insulator voltage distribution value, ambient temperature, humidity and voltage level are used as the original sample set of the random forest model. Multiple training samples are randomly selected from the original sample set with replacement. The sample size each time is 3 / 4 of the original sample, and the sampling is randomly replaced. A regression decision subtree is constructed for each training sample. The output of the decision subtree is the resistance range of each insulator. The convergence error of the parameter calculation model is changed to an error threshold of 0.1%.

8. A zero-value insulator detection system based on voltage distribution characteristics, the system being a system used in a zero-value insulator detection method based on voltage distribution characteristics as claimed in any one of claims 1 to 7, comprising: Transmission line module, including inspection drone unit, tower unit, conductor unit, cross arm unit and insulator unit; A drone inspection module, comprising a drone probe unit and an equivalent capacitance unit, wherein the drone probe unit is used to detect the voltage of an insulator string; The insulator low zero value detection module includes a decision tree unit and a model diagnosis unit, which is used to diagnose whether the insulator string is a low zero value.

9. A terminal comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to execute the steps of the method according to any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images