Power transmission line inspection unmanned aerial vehicle electromagnetic interference shielding case design method based on PSO-BP hybrid algorithm
By optimizing the structural parameters of the electromagnetic interference shielding cover for UAVs based on the PSO-BP hybrid algorithm, the problem of electromagnetic interference during live-line operation of UAVs was solved, achieving a highly efficient electromagnetic shielding effect suitable for engineering applications.
Patent Information
- Application Number
- CN202410967523.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-18
- Publication Date
- 2024-11-22
AI Technical Summary
Existing drones lack effective electromagnetic shielding measures when performing live-line work, which causes electromagnetic interference from the wires to the drones, affecting their normal operation, especially in close-range or contact-based inspection and operation tasks.
An electromagnetic interference shielding cover for UAVs was designed using a PSO-BP hybrid algorithm. By establishing a UAV simulation model and a shielding cover model, and combining the particle swarm optimization algorithm and backpropagation neural network, the structural parameters of the shielding cover were optimized to reduce internal electromagnetic interference and ensure the safe operation of the UAV.
It improves the anti-interference capability of drones, reduces the amount of computation and time, provides a highly efficient shielding effect, and is suitable for practical engineering applications.
Smart Images

Figure CN119004950A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of electromagnetic compatibility in unmanned aerial vehicle inspection of power transmission lines, and specifically relates to a design method for an electromagnetic interference shielding cover of a power transmission line inspection unmanned aerial vehicle based on a PSO-BP hybrid algorithm. Background Art
[0002] The application of drones in power systems is usually to inspect and repair transmission lines, for example: drones set routes for safety inspections, clear ice on transmission lines, and clear debris on lines or towers. Depending on the application, the devices carried by drones are also different. For example, when inspecting, drones will be equipped with high-definition cameras or infrared devices, which can conduct a comprehensive inspection of the line conditions, and import the conditions of towers, lines and even the surrounding environment into the computer through infrared settings for display. In winter, when the temperature is low and there is ice on the line, metal bars can be hung on the arms of the drone to hit the wires to make the ice fall off while ensuring balance. Drones can also be further combined with AI algorithms to automatically inspect according to pre-designed routes and automatically identify abnormal conditions in the line, such as damaged and aging insulators, damaged wires, and debris attached to the line. When drones carry out the above work, they generally need to be close to the transmission line, but at this time the wires will cause certain radio interference to the drones and affect the normal operation of the drones. Therefore, appropriate electromagnetic shielding methods should be used to improve their anti-interference capabilities, so that drones can perform live working tasks well and eliminate hidden dangers in the line.
[0003] There is extensive research on power UAVs at home and abroad, but most of the research focuses on obstacle avoidance, automatic patrol routes, fault identification, etc. Regarding the safety of live operations and flight of UAVs, the main safety measure is to keep a certain distance between the UAV and the power transmission lines. The electromagnetic shielding of the UAV itself is not considered to further improve its ability to resist electromagnetic interference.
[0004] Reference [1]: Dequan Chen, Xianghua Guo, Po Huang, et al. Safety distance analysis of 500kV transmission line tower UAV patrol inspection [J]. IEEE Letters on Electromagnetic Compatibility Practice and Applications, 2020, 2 (4): 124-128. The electric field and magnetic field of the UAV fuselage at different directions from the line are calculated by simulation, and the electromagnetic shielding safety distance of the UAV is determined according to the protection threshold of the electromagnetic field. This method uses the attenuation characteristics of the electric field and magnetic field with distance to perform passive protection using a protective safety distance, which cannot be applied to contact or close-range UAV auxiliary inspection and operation tasks.
[0005] Reference [2]: Wang Yongsheng, Li Wei, Guo Wenqing. Electromagnetic protection technology for drones in strong electromagnetic environments [J]. Safety and Electromagnetic Compatibility, 2020(5):95-99. The electromagnetic coupling path and damage effect of drones under strong electromagnetic radiation were analyzed, and methods for applying shielding grids, high-performance conductive films and hole-type shielding covers were given for the optoelectronic system, antenna port and body of the drone, respectively. However, this research only remained in the theoretical stage and its feasibility was not verified. This method did not optimize the design of the hole-type shielding cover of the body part, and the hole opening method would affect the quality, heat dissipation, communication and other issues of the shielding cover.
[0006] Reference [3]: Cheng Erwei, Chen Yazhou, Liu Weidong, et al. Testing method for shielding effectiveness of UAV shell [J]. High Power Laser and Particle Beams, 2017, 29(11): 55-60. The effect of electromagnetic field distribution inside the UAV body in the direction of electromagnetic wave radiation was calculated by numerical simulation, and the law of shielding effectiveness of the UAV shielding cover changing with the conductivity and magnetic permeability of the shielding material was obtained. This method does not consider the influence of the heat dissipation openings of the UAV shell on its shielding effect, and the openings may cause a large change in the magnetic field. The present invention considers the influence of different heat dissipation opening structures on the shielding effect of the UAV shielding cover, and optimizes the structure of the shielding cover holes to solve the design scheme that meets the UAV anti-interference threshold.
[0007] Reference [4]: Du Jinqiao, Zhang Shiling, Li Naiyi, et al. Calculation of electric field distribution of UHV AC pot-type insulator and optimization of shielding cover structure [J]. High Voltage Technology, 2013, 39(12): 3037-3043. The influence of the shielding cover on the electric field distribution on the surface of the AC pot-type insulator was calculated by finite element method, the design variables of the shielding cover were fitted, and the shielding cover structure was optimized by genetic algorithm to solve and calculate the design scheme with the best shielding effect. This method is only applicable to the electric field shielding of UHV AC pot-type insulators, and is not applicable to the optimization of UAV anti-interference shielding cover, because the electromagnetic radiation to which the UAV is subjected not only has power frequency electric field and power frequency magnetic field, but also has high frequency electromagnetic wave interference.
[0008] Among the existing UAV anti-interference methods, most of the anti-electromagnetic interference research starts from the design of the internal electronic components of the UAV, and less is done from the external shielding cover. Summary of the invention
[0009] The present invention provides a method for designing an electromagnetic interference shielding cover for a power transmission line inspection drone based on a PSO-BP hybrid algorithm. The method mainly considers designing and optimizing the shielding cover of the power drone to improve its anti-interference ability. It has the advantages of strong applicability, small calculation amount, high efficiency and accuracy.
[0010] The technical solution adopted by the present invention is:
[0011] The design method of electromagnetic interference shielding cover for transmission line inspection drone based on PSO-BP hybrid algorithm includes the following steps:
[0012] Step S1: Establish a UAV simulation model and a shielding cover model, and determine the influencing factors of the electromagnetic intensity inside the UAV through the control variable method and sensitivity analysis method;
[0013] Step S2: using the orthogonal experimental design method to determine the combination conditions of different electromagnetic intensity influencing factors, and using the moment method to calculate the internal electric field strength of the UAV under various combination conditions;
[0014] Step S3: Use the particle swarm algorithm PSO with strong global search capability to determine the initial weights and thresholds of the BP neural network, explore the search space of the BP neural network weights, and find the optimal solution by simulating the search of particles in the solution space and performing iterative updates.
[0015] Step S4: Using a BP neural network with strong local optimization capability and back propagation, the weights of the BP neural network are locally adjusted during each iteration to gradually optimize the performance of the network, thereby achieving the acquisition of the optimal solution of the shielding cover structure.
[0016] In the step S1, a UAV simulation model is established by using a three-dimensional visualization entity simulation software, and the internal field strength of the UAV before shielding is solved by FEKO simulation; a shielding cover model is established by using a three-dimensional visualization entity simulation software, and the internal field strength of the UAV after shielding is solved by FEKO simulation;
[0017] In step S1, the excitation source is set to a plane wave, the solution angle is 0 to 360° on the solution surface, a solution direction is set every 18°, the amplitude is 442.3V / m, the frequency is 2.5-3MHz, the solution field is set to a plane passing through the center point of the drone in the horizontal direction, and the observation point is the center point inside the drone with a shielding cover.
[0018] In step S1, the quality of the shielding cover is mainly judged by the size of the interference field strength measured at the center point inside the drone. The smaller the interference field strength, the better the performance of the shielding cover. The factors affecting the electromagnetic strength inside the drone mainly include the opening structure, the material of the shielding cover, and the position of the openings of the shielding cover. By changing the opening structure of the shielding cover (square and round), the material of the shielding cover (aluminum metal, nanocrystals and nickel-iron alloy), and the position of the openings of the shielding cover (the distance from the edge opening to the side length of the drone shielding cover is d1, the distance between the upper and lower square openings is d2, and the number of openings is n), the factors affecting the electromagnetic strength inside the drone can be determined.
[0019] In step S2, when the frequency is 2.5-3 MHz, the internal field strength of the shielding cover with circular opening is much higher than that of the shielding cover with square opening. Therefore, it can be seen that the shielding performance of the shielding cover with square opening is better than that of the shielding cover with circular opening.
[0020] Aluminum shielding covers have the best effect, more stable performance, and relatively low material costs;
[0021] The internal electric field strength of the UAV after shielding and the structural parameters of the shielding cover: the distance between the edge opening and the side length of the UAV shielding cover is d1, the distance between the upper and lower square openings is d2, and the number of openings n is a nonlinear mapping relationship. Under the conditions of satisfying the overall length L, width W, and height H of the shielding cover, as shown in Figure 4(b), d1 and d2 under different schemes are taken as 10, 15, 20, and 25 mm respectively, and the number of openings n is taken as 60, 100, 140, and 180. A three-factor four-level scheme is used to provide learning and training data, and the results of the three-factor four-level scheme are calculated by the moment method.
[0022] The results of the three-factor four-level scheme are calculated by the moment method, and the specific results are shown in Table 1. As can be seen from Table 1, there is a correlation between the electric field strength inside the drone and the distance d1 between the edge opening and the side length of the drone shield, the distance d2 between the upper and lower square openings, and the number of openings n.
[0023] The step S3 comprises the following steps:
[0024] S3.1: Determine the number of input layer nodes, hidden layer nodes, and output nodes of the BP neural network, initialize the connection weights and threshold length of the BP neural network as the encoding of the particle swarm algorithm PSO.
[0025] S3.2: Initialize the population size, maximum number of iterations T, learning factor, particle position, maximum and minimum particle velocity of the particle swarm.
[0026] S3.3: Calculate the fitness function F(x) of the particle to obtain the optimal solution of the particle and the optimal solution of the population. Record the best position of the particle. The process of the particle swarm algorithm PSO selecting the global optimal value is the process of obtaining the optimal weight and threshold of the BP neural network.
[0027] S3.4: If the number of iterations of the particle swarm exceeds the maximum number of iterations T, or the accuracy of its output reaches the expected standard, the particle swarm will stop iterating. After the iteration stops, the particle swarm algorithm PSO assigns the searched optimal position of the population to the initial threshold and weight of the BP neural network. Otherwise, continue S3.3 and continue iterating until the algorithm converges.
[0028] In step S4, the BP neural network training process includes the following steps:
[0029] Step 1: Initialize the BP neural network. According to the number of inputs, outputs, hidden layers and neurons in the hidden layers, initialize the weights between layers, the thresholds a and b between layers and the transfer function.
[0030] Step 2: Calculate the output results in the hidden layer according to the input amount and weight W ij , a and the transfer function to obtain the output value H of the hidden layer.
[0031]
[0032] In the formula, n is the number of input quantities, j is the number of nodes in the hidden layer, f is the transfer function from the input layer to the hidden layer, and H j is the output value of the jth node in the hidden layer; W ij is the weight of the input layer and the hidden layer; x i is the i-th input sample; l is the total number of nodes in the hidden layer; a j is the threshold of the jth node in the hidden layer.
[0033] Step 3: Calculate the output result of the output layer, according to the output of the hidden layer and the weight H jw , b and transfer function to obtain the output result O.
[0034]
[0035] Where m is the number of output quantities, and the value of K is 1, 2, ..., m; K is the output value of the Kth node in the output layer; H jw is the weight of the hidden layer and the output layer; j K b represents the Kth hidden layer input value; K is the threshold of the Kth node in the input layer and the hidden layer.
[0036] Step 4: Error calculation, get the network prediction error e according to the output layer O and the expected Y.
[0037] e=Y k -O k
[0038] Where, k is 1, 2, ..., m; Y k Represents the expected value corresponding to the kth group.
[0039] Step5: Update weights and thresholds.
[0040] Determine whether the result of the algorithm meets the error. If not, return to Step 2.
[0041] Step S5: The shielding effect of the shielding cover under the above design scheme is solved by the moment method. Specifically, the UAV with the optimized shielding cover installed is exposed to the industrial frequency electromagnetic field, the maximum electric field strength inside the UAV is calculated, and the internal field strength of the UAV is analyzed to see whether it meets the national standard. In addition, the shielding cover should not only weaken the radio interference caused by the power transmission line to the UAV, but also ensure that the control signal of the UAV is not shielded. The transmitting field strength of the UAV remote control is taken as the amplitude of the plane wave irradiation. The receiving field strength of the UAV antenna before and after the shielding is calculated by FEKO simulation to analyze whether the shielding cover shields the UAV communication signal.
[0042] The present invention provides a method for designing an electromagnetic interference shielding cover for a transmission line inspection drone based on a PSO-BP hybrid algorithm, and the beneficial effects are as follows:
[0043] 1) The present invention adopts the PSO-BP hybrid algorithm to predict the internal interference field strength of the UAV. The PSO-BP hybrid algorithm optimizes the weights and thresholds of the neural network through the global search capability of the particle swarm algorithm. It combines the global search capability of particle swarm optimization (PSO) and the local optimization capability of back propagation (BP), which can be used for weight adjustment and optimization of neural networks. It ensures that the UAV can perform live-line operation tasks safely and effectively, reduces the amount of calculation and calculation time, is convenient and fast, and is conducive to engineering applications.
[0044] 2) Aiming at the optimal design of the anti-interference shielding cover of the electric UAV, the present invention provides a design method of the electromagnetic interference shielding cover of the transmission line inspection UAV based on the PSO-BP hybrid algorithm. The influence of the heat dissipation openings of the shielding cover on its shielding effect is considered, and the multi-dimensional nonlinear function mapping relationship between the shielding cover structure and its shielding effect is obtained through PSO-BP hybrid algorithm training, and then the optimal design scheme of the shielding cover is solved. Compared with the exhaustive method, it saves calculation time and is more suitable for practical engineering applications.
[0045] 3) The present invention uses an optimization algorithm to establish a multi-dimensional nonlinear function mapping relationship between the shielding cover structural parameters and the internal field strength of the UAV, calculates the internal field strength of the UAV under different shielding cover structural parameters, and determines the optimal shielding cover design scheme, which reduces the amount of calculation and calculation time, is convenient and fast, and is conducive to engineering applications. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments:
[0047] Figure 1 It is a schematic diagram of the process of the present invention.
[0048] Figure 2 This is the solution flow chart of the PSO-BP hybrid algorithm.
[0049] Figure 3 Schematic diagram of a three-dimensional simulation model of a drone according to an embodiment of the present invention.
[0050] FIG4( a ) is a schematic diagram of a two-dimensional structure of a shield cover of a drone according to an embodiment of the present invention;
[0051] FIG4( b ) is a three-dimensional simulation model of the shielding cover of the UAV according to an embodiment of the present invention.
[0052] Figure 5 Schematic diagram of the field strength inside the drone under different radiation directions of the plane wave according to an embodiment of the present invention.
[0053] FIG6( a ) is a diagram comparing the shielding effects of the heat dissipation openings of the shielding cover of the drone according to an embodiment of the present invention in the shapes of round, square and strip;
[0054] FIG6( b ) is a diagram comparing the shielding effects of the shielding cover of the drone according to an embodiment of the present invention using three materials: aluminum, nanocrystalline and nickel-iron alloy.
[0055] Figure 7 It is the BP neural network training output of the embodiment of the present invention.
[0056] Figure 8 This is the field intensity pattern inside the shielded drone according to an embodiment of the present invention.
[0057] Fig. 9This is a simulation diagram of the induced current on the surface of the shielded drone at different frequencies according to an embodiment of the present invention. DETAILED DESCRIPTION
[0058] The design method of electromagnetic interference shielding cover for transmission line inspection drone based on PSO-BP hybrid algorithm includes establishing a three-dimensional simulation model of drone and a fine model of shielding cover, determining the influencing factors of electromagnetic intensity inside drone through control variable method and sensitivity analysis method; adopting orthogonal experimental design method to determine the combination conditions of different factors, and using moment method to calculate the internal electric field intensity of drone under various combination conditions; adopting particle swarm algorithm (PSO) with strong global search capability to determine the initial weight and threshold of BP neural network, exploring the search space of neural network weight, and finding the optimal solution by simulating the search of particles in solution space and iterative updating; adopting BP neural network with strong local optimization capability and back propagation to adjust the weight of neural network locally in each iteration process, so as to gradually optimize the performance of the network, and then realize the acquisition of the optimal solution of shielding cover structure.
[0059] like Figure 1 The flowchart of the method of the present invention is as follows: The optimization design method of the anti-interference shielding cover of the electric drone provided by the present invention comprises the following steps:
[0060] S1. Establish a three-dimensional simulation model of the UAV and a detailed model of the shielding cover, and determine the influencing factors of the electromagnetic intensity inside the UAV through the control variable method and sensitivity analysis method;
[0061] S2. Use the orthogonal experimental design method to determine the combination conditions of different factors, and use the moment method to calculate the internal electric field strength of the UAV under various combination conditions;
[0062] S3. Use the particle swarm algorithm (PSO) with strong global search capability to determine the initial weights and thresholds of the BP neural network, explore the search space of the neural network weights, and find the optimal solution by simulating the search of particles in the solution space and performing iterative updates;
[0063] S4. A BP neural network with strong local optimization capability and back propagation is used to locally adjust the weights of the neural network during each iteration to gradually optimize the performance of the network and thereby achieve the optimal solution of the shielding cover structure. The process is as follows: Figure 2 shown.
[0064] S5. Verify the shielding effect of the shielding cover under this solution.
[0065] The specific implementation process includes the following steps:
[0066] Taking a typical 500kV overhead transmission line as an example, the conductors are arranged horizontally, the height of the transmission line to the ground is 30m, split conductors are used, the number of splits is 4, the distance between the split conductors is 40cm, the radius of the conductor is 1.7cm, and the spacing between phases is 10m. By consulting the data, it can be found that the frequency of electromagnetic interference generated by the line is between 0.15 and 10MHz. When the drone is working under power, it is located at the same height as the transmission line and the horizontal distance between the drone and the conductor is 1m. The frequency of the maximum interference field strength is 2.72MHz, and the radio interference field strength is 442.3V / m.
[0067] In order to study the impact of electromagnetic interference generated by power transmission lines on the internal environment of drones, and to specifically reduce the electromagnetic interference of power transmission lines on the inside of drones, the DJI Mavic 2 Industrial Edition, which is commonly used in power grids, was selected as the model for modeling and simulation. This type of drone has a small body, flexible operation, long flight time, and is cost-effective. It can be sent back to the original factory for repair after damage, and it also has an internal obstacle avoidance system. Its body is made of carbon fiber composite materials, which has a compact structure, certain mechanical strength and resistance to drop. In terms of size, when the drone arms are unfolded, the dimensions are 322mm×242mm×84mm, and when the drone arms are folded, the dimensions are 214mm×91mm×84mm. Three-dimensional modeling is performed using Soildworks modeling software, and the model of the physical object is as follows Figure 3 shown.
[0068] In order to prevent the shielding cover of the drone from being too large and the wings from hitting the shielding cover when rotating, the distance between the shielding cover and the drone fuselage is set to 5 mm, and the thickness of the shielding cover is 1 mm. This structure can avoid collisions between the various components. For the sake of heat dissipation, signal reception and weight reduction, openings are set on the outside of the shielding cover. In terms of parameters such as spacing and number of openings, taking square openings as an example, while ensuring that the total opening area remains unchanged, the number of openings is changed, and the distance between the edge opening and the side length of the shielding cover is d1, and the distance between the upper and lower square openings is d2. By changing these two distances and the number of openings, the structural parameters of the shielding cover can be changed. These three parameters are also the parameters that need to be improved in the subsequent optimization design. Figure 4 (a) is a schematic diagram of the various parameters of the shielding cover. Taking the opening shape as a square, the total area of the shielding cover opening is 5000mm 2 Taking a shield cover with 140 openings and 2 rows of openings as an example, a three-dimensional model of the shield cover of a drone is established as shown in FIG4( b ).
[0069] When the drone operates normally without a shield, the interference field strength at all angles is around 15V / m. Figure 5As shown, according to the national standards, in order to ensure the normal operation of the UAV, the internal interference field strength of the light and small UAV should not be higher than 10V / m. Therefore, when the UAV is working 1m close to the power line, its internal electronic components will be affected by the electromagnetic interference generated by the line. Therefore, in order to ensure the normal operation of the UAV, it is necessary to add a shielding cover on the outside to reduce electromagnetic interference.
[0070] In order to facilitate the manufacture of the shielding cover and further reduce the material cost, the shielding cover is provided with an opening, and the shape of the opening is selected to be square and circular. The opening is 140, and the total opening area is 5000mm 2 , d1 is 18cm, d2 is 12cm as an example, and the shielding effects of shielding covers with different opening shapes and different materials are compared. The excitation source is set to the maximum value, and the simulation calculation is performed at the frequency of 2.5, 2.6, 2.7, 2.8, 2.9, and 3MHz. The calculated interference field strength is shown in Figure 6(a). When the frequency is 2.5-3Mhz, the internal field strength of the shielding cover with a circular opening is much higher than that of the shielding cover with a square opening. Therefore, it can be seen that the shielding performance of the square opening is better than that of the circular opening.
[0071] The material comparison is also based on an opening of 140 and a total opening area of 5000cm 2 , taking the shielding cover with d1 of 18cm and d2 of 12cm as an example, aluminum metal, nanocrystalline and nickel-iron alloy materials are added to the material library in the simulation, and added to the surface of the shielding cover, and the thickness of the shielding cover is set to 1mm. The shielding effects of different materials at various frequencies are shown in Figure 6(b). By comprehensively comparing the anti-electromagnetic interference levels of the three materials at various frequencies and the costs of each material, it can be found that the aluminum shielding cover has the best effect, relatively stable performance, and relatively low material cost.
[0072] In terms of factors such as the opening spacing and number of openings of the shielding cover, on the basis of ensuring that the total opening area remains unchanged, the number of openings is changed, that is, the area of a single opening with different numbers of openings is different. Let the distance between the edge opening and the side length of the drone shielding cover be d1, and let the distance between the upper and lower square openings be d2. By changing these two distances and the number of openings, the structural parameters of the shielding cover are changed. Therefore, the optimal design of the shielding cover is actually to find the d1, d2, and n corresponding to the lowest internal electric field strength of the drone under the action of the excitation source. Under the condition of meeting the external dimensions, d1 and d2 under different schemes are taken as 10, 15, 20, and 25 mm respectively, and the number of openings n is taken as 60, 100, 140, and 180. Therefore, a three-factor four-level scheme can be used to provide learning and training data. Through simulation calculations, the three-factor four-level scheme and the calculation results are shown in Table 1:
[0073] Table 1 Orthogonal scheme and calculation results
[0074]
[0075] As shown in Table 1, the change of structural parameters has a great influence on the internal interference field strength. Even if the number of openings is the same, the electric field strength will be greatly different when the positions are different. Sometimes a slight shift in the position of the opening will cause a doubling of the change. The maximum difference between different structures can even reach 85.98%. Therefore, it is necessary to optimize the design of the shielding cover.
[0076] In order to verify the accuracy of the neural network model after particle swarm optimization, the 16 sets of data obtained through simulation are used as sample data for learning and training, and four random sets of data are used as verification data, and the remaining 12 sets of data are used for training. The input layer node is set to 3, and the output layer node is set to 1. Through calculation, when the hidden layer node is 5, the training output of the predicted neural network is as follows Figure 7 As shown in the figure, the minimum mean square error is 0.196, which is the best hidden layer.
[0077] After calculation, it is found that when d1=24.95mm, d2=23.98mm, and n=100, the corresponding electric field strength is the minimum value of 5.699V / m. In order to verify that the shielding effect of the drone can reach this level when the structural shielding cover is installed, a simulation model under this structure is established, and the simulation is carried out under the same conditions. The frequency is 2.72MHz, the excitation source is set to a plane wave, and the irradiation is carried out in a direction of 0 to 360 degrees. The amplitude is set to 442.3V / m, and the observation point is the center point of the drone body. The directional diagram of the field strength inside the drone body is obtained through simulation calculation, as shown in the figure. Figure 8 As shown. Figure 8 It can be seen that the interference field strength in all directions inside the drone is lower than the specified threshold of 10V / m, and its average electric field strength is 4.98V / m, with an error of 12.63%. When the angle is 9°, the interference field strength is the maximum value of 6.202V / m, and when the angle is 99°, the interference field strength is the minimum value of 1.466V / m. The distribution diagram of the induced current on the surface of the drone shield at 9° is extracted in the simulation, as shown in Fig. 9 As shown. Fig. 9 It can be seen that the anti-interference ability of the drone under the shield is strong, the current distributed on the surface of the shield is relatively uniform, and the number of openings does not cause a significant change in the current on the surface of the shield, so the structure of the shield is relatively reasonable. The figure also shows that the current on the surface of a certain wing is high, which is caused by the angle of the incident excitation source being close to the angle of the wing.
Claims
1. Design method of electromagnetic interference shielding cover for transmission line inspection drone based on PSO-BP hybrid algorithm, Features The following steps are involved: Step S1: Establish a UAV simulation model and a shielding cover model to determine the factors affecting the electromagnetic intensity inside the UAV; Step S2: using the orthogonal experimental design method to determine the combination conditions of different electromagnetic intensity influencing factors, and calculating the electric field strength inside the UAV under various combination conditions; Step S3: Use the particle swarm algorithm PSO to determine the initial weights and thresholds of the BP neural network, explore the search space of the BP neural network weights, and find the optimal solution by simulating the search of particles in the solution space and performing iterative updates; Step S4: In each iteration process, the weights of the BP neural network are locally adjusted to gradually optimize the performance of the network, thereby achieving the acquisition of the optimal solution of the shielding cover structure.
2. According to the design method of electromagnetic interference shielding cover of transmission line inspection drone based on PSO-BP hybrid algorithm in claim 1, Features: In the step S1, a UAV simulation model is established by using a three-dimensional visualization entity simulation software, and the internal field strength of the UAV before shielding is solved by FEKO simulation; a shielding cover model is established by using a three-dimensional visualization entity simulation software, and the internal field strength of the UAV after shielding is solved by FEKO simulation.
3. According to the design method of electromagnetic interference shielding cover of transmission line inspection drone based on PSO-BP hybrid algorithm in claim 2, Features: The excitation source is set to a plane wave, the solution angle is 0 to 360° on the solution surface, a solution direction is set every 18°, the solution field is set to a plane passing through the center point of the drone in the horizontal direction, and the observation point is the center point inside the drone with a shield.
4. According to the design method of electromagnetic interference shielding cover of transmission line inspection drone based on PSO-BP hybrid algorithm as claimed in claim 3, Features: The quality of the shielding cover is judged by the interference field strength measured at the center point inside the drone. The smaller the interference field strength, the better the performance of the shielding cover. The factors affecting the electromagnetic intensity inside the drone include the opening structure, the material of the shielding cover, and the position of the opening of the shielding cover. Change the opening structure, material, and position of the shielding cover: the distance between the edge opening and the side length of the drone shielding cover is d 1 , let the distance between the upper and lower square openings be d 2 , the number of openings n; it can determine the factors affecting the electromagnetic strength inside the drone.
5. According to the design method of electromagnetic interference shielding cover of transmission line inspection drone based on PSO-BP hybrid algorithm in claim 1, Features: In the step S2, When the frequency is 2.5-3MHz, the internal field strength of the shielding cover with a circular opening is higher than that of the shielding cover with a square opening. Therefore, it can be seen that the shielding performance of the shielding cover with a square opening is better than that of the shielding cover with a circular opening. The internal electric field strength of the UAV after shielding and the structural parameters of the shielding cover: The distance between the edge opening and the side length of the UAV shielding cover is d 1 , let the distance between the upper and lower square openings be d 2 , the number of openings n is a nonlinear mapping relationship; under the condition that the overall length L, width W, and height H of the shielding cover are met, the d under different schemes 1 d 2 The values are 10, 15, 20, and 25 mm respectively, and the number of openings n is determined to be 60, 100, 140, and 180. A three-factor four-level scheme is used to provide learning and training data, and the results of the three-factor four-level scheme are calculated by the moment method.
6. According to the design method of electromagnetic interference shielding cover of transmission line inspection drone based on PSO-BP hybrid algorithm in claim 5, Features: The results of the three-factor four-level solution calculated by the moment method show that the electric field strength inside the drone is equal to the distance d between the edge opening and the side length of the drone shield. 1 , the distance between the upper and lower square openings is d 2 , there is a correlation between the number of openings n.
7. According to claim 1, the design method of electromagnetic interference shielding cover for transmission line inspection drone based on PSO-BP hybrid algorithm, Features: The step S3 comprises the following steps: S3.1: Determine the number of input layer nodes, hidden layer nodes, and output nodes of the BP neural network, initialize the connection weights and threshold length of the BP neural network, and use them as the encoding of the particle swarm algorithm PSO; S3.2: Initialize the population size, maximum number of iterations T, learning factor, particle position, maximum and minimum values of particle velocity of the particle swarm; S3.3: Calculate the fitness function F(x) of the particle to obtain the optimal solution of the particle and the optimal solution of the population; record the best position of the particle. The process of the particle swarm algorithm PSO selecting the global optimal value is the process of obtaining the optimal weight and threshold of the BP neural network; S3.4: If the number of iterations of the particle swarm exceeds the maximum number of iterations T, or the accuracy of its output reaches the expected standard, the particle swarm will stop iterating; after the iteration stops, the particle swarm algorithm PSO assigns the initial threshold and weight of the BP neural network to the optimal position of the searched population; otherwise, continue S3.3 and continue to iterate until the algorithm converges.
8. According to the PSO-BP hybrid algorithm based design method of electromagnetic interference shielding cover for power line inspection drones, Features: In step S4, the BP neural network training process includes the following steps: Step 1: Initialize the BP neural network, and initialize the weights between each layer, the thresholds a and b between each layer, and the transfer function according to the number of input quantities, the number of output quantities, the number of hidden layers, and the number of neurons in the hidden layers; Step 2: Calculate the output results in the hidden layer according to the input amount and weight W ij , a and transfer function to obtain the output value H of the hidden layer; In the formula, n is the number of input quantities, j is the number of nodes in the hidden layer, f is the transfer function from the input layer to the hidden layer, and H j is the output value of the jth node in the hidden layer; W ij is the weight of the input layer and the hidden layer; x i is the i-th input sample; l is the total number of nodes in the hidden layer; a j is the threshold of the jth node in the hidden layer; Step 3: Calculate the output result of the output layer, according to the output of the hidden layer and the weight H jw , b and transfer function to obtain the output result O; Where m is the number of output quantities, and the value of K is 1, 2, ..., m; K is the output value of the Kth node in the output layer; H jw is the weight of the hidden layer and the output layer; j K b represents the Kth hidden layer input value; K is the threshold of the Kth node in the input layer and hidden layer; Step 4: Error calculation, get the network prediction error e according to the output layer O and the expected Y; e=Y k -EITHER k Where, k is 1, 2, ..., m; Y k represents the expected value corresponding to the kth group; Step 5: Update weights and thresholds; Determine whether the result of the algorithm meets the error. If not, return to Step 2.
9. According to the method for designing electromagnetic interference shielding cover of transmission line inspection drone based on PSO-BP hybrid algorithm in claim 1, Features: The step S5 is also included: solving the shielding effect of the shielding cover under the above design scheme by using the moment method, specifically: Expose the drone with the optimized shield to industrial frequency electromagnetic field, calculate the maximum electric field strength inside the drone, and analyze whether the field strength inside the drone meets the national standards; In addition, while the shielding cover weakens the radio interference caused by the transmission line to the UAV, it should also ensure that the control signal of the UAV is not shielded. The transmission field strength of the UAV remote control is taken as the amplitude of the plane wave radiation. The receiving field strength of the UAV antenna before and after shielding is calculated through FEKO simulation to analyze whether the shielding cover shields the UAV communication signal.
Citation Information
Cited By
Discontinuous conducting layer equivalent conductivity test system and method based on simulation inversion and medium
CN121978418A
