An integrated three-dimensional inspection method for space, ground and space based on multi-agent reinforcement learning

Abstract

The present invention belongs to the technical field of three-dimensional inspection methods, and in particular, to an integrated three-dimensional inspection method of air, space and land based on multi-intelligent voxel reinforcement learning, including the following steps: (S1) modeling of a three-dimensional inspection system for a substation; (S2) calculating the pitch angle, yaw angle, and rotation angle of ground mobile inspection equipment for subsequent calibration; (S3) calculating the pitch angle, yaw angle, and rotation angle of aerial mobile inspection equipment for subsequent calibration; (S4) describing the communication mechanism between ground mobile inspection equipment, aerial mobile inspection equipment, and ground fixed inspection equipment; (S5) optimizing the performance of ground mobile inspection equipment and aerial mobile inspection equipment using a multi-intelligent voxel reinforcement learning algorithm to realize digital and intelligent inspection. The integrated three-dimensional inspection method of air, space and land based on multi-intelligent voxel reinforcement learning of the present invention solves the problems of time-consuming and labor-intensive traditional substation inspections, low precision, and low intelligence level, and realizes digital and intelligent inspections.

Description

Technical Field

[0001] The present invention belongs to the technical field of three-dimensional inspection methods, and in particular to an air-ground-ground integrated three-dimensional inspection method based on multi-intelligent voxel reinforcement learning. Background Art

[0002] As a key part of the modern power system, the safety of the substation plays a key role in the modern power supply system. Since the substation equipment is in uninterrupted operation for a long time, it is susceptible to sudden lightning, heavy rain and other harsh environmental factors, which may cause various faults in the power equipment or even power outages. Therefore, it is very important to ensure the normal operation of the transformer, busbar, circuit breaker and other related equipment inside the substation, which will not only help to ensure the supply efficiency, but also effectively avoid risks and reduce unnecessary impacts.

[0003] The traditional substation inspection method that relies on manual labor or a combination of inspection equipment and manual labor has the following disadvantages: (1) It fails to make good use of communication technology. The artificial intelligence algorithm makes the inspection time long. Since there are many devices in the substation, the number of inspections required is large, and potential equipment failures cannot be detected in time; (2) The accuracy is not high. Manual inspections often cause fatigue to personnel and affect the inspection accuracy; the safety factor is low and cannot meet and adapt to the needs of modern power inspections.

[0004] The Chinese invention patent application with publication number CN108764742A discloses a state management method, device and equipment for a substation inspection robot. The use of robots to replace manual inspections saves labor intensity. However, after all, robot inspections can only be carried out on the ground, and the intelligent body inspection elements are single. It is impossible to detect faults in all locations in a timely manner, which will affect the accuracy of the inspection results. Summary of the invention

[0005] Technical purpose: In order to overcome the deficiencies in the prior art, the present invention provides an integrated three-dimensional inspection method of air, space and land based on multi-intelligent voxel reinforcement learning.

[0006] Technical solution: To achieve the above purpose, the present invention discloses a three-dimensional inspection method for space-ground integration based on multi-intelligent voxel reinforcement learning, which includes the following steps:

[0007] (S1) Modeling of substation three-dimensional inspection system;

[0008] (S2) Calculating the pitch angle, yaw angle, and rotation angle of the ground mobile inspection equipment for subsequent calibration;

[0009] (S3) Calculating the pitch angle, yaw angle, and rotation angle of the aerial mobile inspection equipment for subsequent calibration;

[0010] (S4) Explain the communication mechanism between ground mobile inspection equipment, aerial mobile inspection equipment, and ground fixed inspection equipment;

[0011] (S5) Use multi-intelligent element reinforcement learning algorithm to optimize the performance of ground mobile inspection equipment and aerial mobile inspection equipment to realize digital and intelligent inspection.

[0012] Preferably, step (S2) is specifically divided into the following steps:

[0013] (A1) Definition There are three parallel electric poles in the substation. is its projection on the image plane, is the optical center position of the camera, respectively , Establish the camera coordinate system of the inspection equipment for the center of the circle With the world coordinate system ;

[0014] (A2) Definition They are the elevation angle, rotation angle of the camera and the yaw angle of the ground mobile inspection equipment. is the height of the ground mobile inspection equipment, any point In the camera coordinate system The normalized expression is:

[0015] ;

[0016] (A3) The relationship between the camera coordinate system and the captured image coordinate system is calculated as follows:

[0017] , ,in, is a constant coefficient;

[0018] (A4) When the world coordinate system and The relationship between the image coordinates and the camera angle when the axis is 0 is expressed as:

[0019] , ,

[0020] When the world coordinate system and The relationship between the image coordinates and the camera angle when the axis is 0 is expressed as: ;

[0021] (A5) The camera's pitch angle, yaw angle, and rotation angle are related to The function of .

[0022] Preferably, step (S3) is specifically divided into the following steps:

[0023] (B1) The pitch angle, roll angle and yaw angle of the aerial mobile inspection equipment are defined as , , The coordinate system is established in the following way: the positive direction of the pitch axis points to the right side of the aerial mobile inspection device, the positive direction of the roll axis points to the front of the aerial mobile inspection device, and the positive direction of the yaw axis is perpendicular to the bottom of the aerial mobile inspection device;

[0024] (B2) Calculate the total lift of the aerial mobile inspection equipment F It is expressed as: ,

[0025] in, Indicates the number of motors in the aerial mobile inspection equipment. is a constant coefficient affected by hardware materials, is the air density, is the motor area, Respectively The propeller radius and rotational acceleration of each motor;

[0026] (B3) The corresponding lift and , , The calculation formulas are:

[0027] ,

[0028] ,

[0029] ,

[0030] in, Indicates the distance between the motor and the center of gravity of the aerial mobile inspection equipment;

[0031] (B4) Based on (A3), the pitch angle, roll angle and yaw angle of the aerial mobile inspection equipment are calculated as functions of lift: ;

[0032] (B5) Calculate the distance between the smart sensor and the aerial mobile inspection equipment , , , are the horizontal distance and the vertical distance respectively, where The straight-line distance between the sensor and the obstacle. is the azimuth between the obstacle and the aerial mobile inspection equipment;

[0033] (B6) Calculate the success rate of aerial mobile inspection equipment in identifying a target as follows:

[0034] .

[0035] Preferably, step (S4) is specifically divided into the following steps:

[0036] (C1) Assume that the distance between any two of the aerial mobile inspection equipment, the ground mobile inspection equipment, and the ground fixed inspection equipment is , the azimuth is And its communication includes line-of-sight transmission and non-line-of-sight transmission, and the probability of the two types of communication is:

[0037] ,

[0038] in represents the probability of line-of-sight transmission, represents the non-line-of-sight transmission probability, is the signal steering vector, is a constant affected by the environment;

[0039] (C2) Calculate the communication channel gain as:

[0040] ,

[0041] In the above formula, is the discount factor, is the speed of light, is the communication frequency, is the channel fading factor;

[0042] (C3) The formula for calculating the communication rate between any two is: ,

[0043] in,

[0044] ,

[0045] ,

[0046] ,

[0047] In the above formula, , is the communication rate between the aerial mobile inspection equipment and the ground mobile inspection equipment, is the communication rate between the aerial mobile inspection equipment and the ground fixed inspection equipment. It is the communication rate between the ground mobile inspection equipment and the ground fixed inspection equipment; is the signal transmission power, For noise, for The inverse of a function, is the block error probability caused by short data packet transmission, V is the channel divergence function, SNR is the signal-to-noise ratio, M is the block length, B is the channel width, t is the time coefficient, For each time slot duration;

[0048] (C4) Communication delay is:

[0049] ,

[0050] is the channel bandwidth, The number of pixel bits used to transmit video or pictures. is the video or image resolution, The number of devices communicating.

[0051] Preferably, step (S5) is specifically divided into the following steps:

[0052] (D1) Define the ground mobile inspection equipment agent The state space and action space are: , ; Aerial mobile inspection equipment intelligent agent The state space and action space are: , ; The state space and action space of the ground fixed inspection equipment meta-agent are: , ;

[0053] (D2) Define the ground mobile inspection agent In Status Next action The reward function is:

[0054] ,

[0055] in, is a constant factor, the reward maximization is expressed as:

[0056] ,

[0057] in, is the total number of ground mobile inspection agents, is the discount factor;

[0058] (D3) Ground Mobile Inspection Agent In state The state value function of the strategy under is:

[0059] ,

[0060] in, For the agent in state action The policy function, is the strategy parameter, based on the above ground mobile inspection agent Choose the best strategy;

[0061] (D4) Define the Aerial Mobile Inspection Equipment Agent In Status Next action The reward function is:

[0062] ,

[0063] in Angle Regarding the three lift functions, The probability of identifying a target for aerial mobile inspection equipment;

[0064] Reward maximization is expressed as: ,

[0065] in is the total number of aerial mobile inspection agents, is the discount factor;

[0066] (D5) Aerial Mobile Inspection Agent In state The state value function of the strategy under is:

[0067] ,

[0068] in For the agent in state action The policy function, is the strategy parameter. The above formula is used to evaluate the performance of the action under the state. Therefore, the aerial mobile inspection equipment agent Choosing the best strategy , the corresponding optimal value function is:

[0069] .

[0070] D6) Assume that the observation results of all ground mobile inspections and aerial mobile inspections local agents are The observations are sent to the ground fixed inspection equipment. The meta-agent continuously updates the parameters through meta-learning. Get the optimal strategy;

[0071] (D7) Strategies for ground-based fixed inspection equipment The advantage function is:

[0072] ,

[0073] in , is the state transition probability;

[0074] (D8) The loss function is defined as: ,

[0075] in ;

[0076] (D9) The policy gradient of the loss function is: ,

[0077] in , ;

[0078] (D10) The parameter update strategy for ground fixed inspection equipment is: , is the meta-learning rate;

[0079] (D11) The parameters of the ground fixed inspection equipment are further transmitted to the local agents of the ground mobile inspection equipment and the aerial mobile inspection equipment so that they can update their own optimal strategies.

[0080] The present invention has the following technical effects:

[0081] (1) The present invention utilizes ground mobile inspection equipment, aerial mobile inspection equipment and ground fixed inspection equipment to realize the acquisition of substation video images that integrate air, space and land. The multi-intelligent voxel reinforcement learning algorithm optimizes the data transmission between aerial mobile inspection equipment, ground mobile inspection equipment and ground fixed inspection equipment, and the ground fixed inspection equipment analyzes the data to modulate the relevant operating angles of the ground mobile inspection equipment camera and the aerial mobile inspection equipment, thereby solving the problems of traditional substation inspection being time-consuming and labor-intensive, low in precision and low in intelligence level, and realizing the digitalization of inspection.

[0082] (2) This solution aims to solve the complex substation scenario. By extracting the line features of the environment and establishing a ground inspection space coordinate system, the pitch angle, yaw angle and rotation angle of the ground mobile inspection equipment are calibrated. The pitch angle, roll angle, yaw angle and target successful recognition probability of the aerial mobile inspection equipment are calculated according to the principle of dynamics. Among them, the ground, aerial mobile inspection equipment and the ground fixed inspection equipment are connected through short data packets to achieve low-latency data transmission and use multi-agent reinforcement learning algorithms. While the local agent (ground and aerial mobile inspection equipment) is making decisions and learning, its data is used as the input of the meta-agent (ground fixed inspection equipment) to optimize its parameters and feed back the optimized parameters to the local agent, thereby realizing the integrated coordinated operation of air, space and ground, and meeting the requirements of digital substation inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] Figure 1 This is a schematic diagram of the spatial coordinate system of ground mobile inspection equipment. DETAILED DESCRIPTION

[0084] The following is combined with Figure 1 The principles and features of the present invention are described, and the examples given are only used to explain the present invention, rather than to limit the scope of the present invention.

[0085] (S1) Modeling of the three-dimensional inspection system of substations; intelligent inspection of substations adopts the coordinated operation of ground mobile inspection equipment, aerial mobile inspection equipment and ground fixed inspection equipment to integrate air-ground-air three-dimensional inspection; during the operation, the mobile inspection equipment is affected by the internal and external environment and may collide, yaw, etc. Therefore, it is necessary to calibrate the internal and external parameters of the mobile inspection equipment to improve the quality of data collection, and the multi-intelligent element reinforcement learning algorithm is used to optimize the system parameters.

[0086] (S2) Calculate the pitch angle, yaw angle, and rotation angle of the ground mobile inspection equipment for subsequent calibration. Use the detector to extract the line features of the image, so as to use the straight line information to describe the characteristics of the inspection equipment camera to the external environment, which specifically includes the following steps:

[0087] (A1) Definition There are three parallel electric poles in the substation. is its projection on the image plane, is the optical center position of the camera, respectively , The center of the circle establishes the camera coordinate system carried by the inspection equipment With the world coordinate system ;

[0088] (A2) Definition They are the elevation angle, rotation angle of the camera and the yaw angle of the ground mobile inspection equipment. is the height of the ground mobile inspection equipment. The Euclidean transformation is used to realize the conversion between the world coordinate system and the camera coordinate system. In the camera coordinate system The normalized expression is:

[0089] ;

[0090] (A3) The relationship between the camera coordinate system and the captured image coordinate system is calculated as follows:

[0091] , ,in, is a constant coefficient;

[0092] (A4) Due to the camera and world coordinate system , The axes correspond to the image coordinate system , Axis, from which the world coordinate system is calculated and The relationship between the image coordinates and the camera angle when the axis is 0 is expressed as:

[0093] , ,

[0094] World Coordinate System and The relationship between the image coordinates and the camera angle when the axis is 0 is expressed as:

[0095] ;

[0096] (A5) According to the calculation of the camera's pitch angle, yaw angle, and rotation angle, The function of .

[0097] (S3) Calculate the pitch angle, yaw angle, and rotation angle of the aerial mobile inspection equipment for subsequent calibration. The aerial mobile inspection equipment is equipped with intelligent sensors to sense obstacles, flying insects and other interference factors, so as to plan the flight trajectory and intelligently avoid external interference. It is also equipped with a camera, and the adjustment principle of the camera is the same as that of the ground mobile inspection equipment. Specifically, it includes the following steps:

[0098] (B1) The pitch angle, roll angle and yaw angle of the aerial mobile inspection equipment are defined as , , The coordinate system is established in the following way: the positive direction of the pitch axis points to the right side of the aerial mobile inspection device, the positive direction of the roll axis points to the front of the aerial mobile inspection device, and the positive direction of the yaw axis is perpendicular to the bottom of the aerial mobile inspection device;

[0099] (B2) Calculate the total lift of the aerial mobile inspection equipment It is expressed as:

[0100] ,

[0101] in, Indicates the number of motors in the aerial mobile inspection equipment. Constant coefficients affected by hardware materials, is the air density, is the motor area, Respectively The propeller radius and rotational acceleration of each motor;

[0102] (B3) The corresponding lift and , , The calculation formulas are:

[0103] ,

[0104] ,

[0105] ,

[0106] in, Indicates the distance between the motor and the center of gravity of the aerial mobile inspection equipment;

[0107] (B4) Based on (A3), the pitch angle, roll angle and yaw angle of the aerial mobile inspection equipment are calculated as functions of lift: ;

[0108] (B5) Intelligent sensors are used to detect targets and calculate their distance from aerial mobile inspection equipment. , , , are the horizontal distance and the vertical distance respectively, where The straight-line distance between the sensor and the obstacle. is the azimuth between the obstacle and the aerial mobile inspection equipment;

[0109] (B6) Calculate the success rate of aerial mobile inspection equipment in identifying a target as follows:

[0110] .

[0111] (S4) Explain the communication mechanism between ground mobile inspection equipment, aerial mobile inspection equipment, and ground fixed inspection equipment;

[0112] (C1) Assume that the distance between any two of the aerial mobile inspection equipment, the ground mobile inspection equipment, and the ground fixed inspection equipment is , the azimuth is , and its communication includes line-of-sight transmission and non-line-of-sight transmission, the probability of the two types of communication is:

[0113] ,

[0114] in represents the probability of line-of-sight transmission, represents the non-line-of-sight transmission probability, is the signal steering vector, is a constant affected by the environment;

[0115] (C2) Calculate the communication channel gain as:

[0116] ,

[0117] In the above formula, is the discount factor, is the speed of light, is the communication frequency, is the channel fading factor;

[0118] (C3) The formula for calculating the communication rate between any two is:

[0119] ,

[0120] in,

[0121] ,

[0122] ,

[0123] ,

[0124] In the above formula, , is the communication rate between the aerial mobile inspection equipment and the ground mobile inspection equipment, is the communication rate between the aerial mobile inspection equipment and the ground fixed inspection equipment. It is the communication rate between the ground mobile inspection equipment and the ground fixed inspection equipment; is the signal transmission power, For noise, for The inverse of a function, is the block error probability caused by short data packet transmission, V is the channel divergence function, SNR is the signal-to-noise ratio, M is the block length, B is the channel width,t is the time coefficient, For each time slot duration;

[0125] (C4) Communication delay is:

[0126] ,

[0127] is the channel bandwidth, The number of pixel bits used to transmit video or pictures. is the video or image resolution, The number of devices communicating.

[0128] (S5) Use multi-intelligent element reinforcement learning algorithm to optimize the performance of ground mobile inspection equipment and aerial mobile inspection equipment to realize digital and intelligent inspection.

[0129] (D1) The multi-agent meta-reinforcement learning algorithm of this scheme has three agents: aerial mobile inspection local agent, ground mobile inspection local agent, and ground fixed inspection equipment meta-agent. Definition of ground mobile inspection equipment agent The state space and action space are: , ; Aerial mobile inspection equipment intelligent agent The state space and action space are: , ; The state space and action space of the ground fixed inspection equipment meta-agent are: , ;

[0130] (D2) Define the ground mobile inspection agent In Status Next action The reward function is:

[0131] ,

[0132] in, is a constant factor, the reward maximization is expressed as:

[0133] ,

[0134] in, is the total number of ground mobile inspection agents, is the discount factor;

[0135] (D3) Ground Mobile Inspection Agent In state The state value function of the strategy under is: ,

[0136] in, For the agent in state action The policy function, is the strategy parameter, based on the above ground mobile inspection agent Choose the best strategy;

[0137] (D4) Define the Aerial Mobile Inspection Equipment Agent In Status Next action The reward function is:

[0138] ,

[0139] in Angle Regarding the three lift functions, The probability of identifying a target for aerial mobile inspection equipment;

[0140] Reward maximization is expressed as: ,

[0141] in is the total number of aerial mobile inspection agents, is the discount factor;

[0142] (D5) Aerial Mobile Inspection Agent In state The state value function of the strategy under is: ,

[0143] in For the agent in state action The policy function, is the strategy parameter. The above formula is used to evaluate the performance of the action under the state. Therefore, the aerial mobile inspection equipment agent Choosing the best strategy , the corresponding optimal value function is:

[0144] ,

[0145] (D6) Assume that the observation results of all local agents of ground mobile inspection and aerial mobile inspection are The observations are sent to the ground fixed inspection equipment. The meta-agent continuously updates the parameters through meta-learning. Get the optimal strategy;

[0146] (D7) Strategies for ground-based fixed inspection equipment The advantage function is: ,

[0147] in , is the state transition probability;

[0148] (D8) The loss function is defined as: ,

[0149] in ;

[0150] (D9) The policy gradient of the loss function is: ,

[0151] in , ;

[0152] (D10) The parameter update strategy for ground fixed inspection equipment is: , is the meta-learning rate;

[0153] (D11) The parameters of the ground fixed inspection equipment are further transmitted to the local agents of the ground mobile inspection equipment and the aerial mobile inspection equipment so that they can update their own optimal strategies.

[0154] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An integrated three-dimensional inspection method for air, space and ground based on multi-intelligent voxel reinforcement learning, characterized in that: The following steps are involved: (S1) Modeling of substation three-dimensional inspection system; (S2) Calculating the pitch angle, yaw angle, and rotation angle of the ground mobile inspection equipment for subsequent calibration; (S3) Calculating the pitch angle, yaw angle, and rotation angle of the aerial mobile inspection equipment for subsequent calibration; (S4) Explain the communication mechanism between ground mobile inspection equipment, aerial mobile inspection equipment, and ground fixed inspection equipment; (S5) Use multi-intelligent element reinforcement learning algorithm to optimize the performance of ground mobile inspection equipment and aerial mobile inspection equipment to realize digital and intelligent inspection; The step (S2) is specifically divided into the following steps: (A1) Definition There are three parallel electric poles in the substation. is its projection on the image plane, A is the optical center position of the camera, respectively A , Establish the camera coordinate system of the inspection equipment for the center of the circle With the world coordinate system ; (A2) Definition They are the elevation angle, rotation angle of the camera and the yaw angle of the ground mobile inspection equipment. is the height of the ground mobile inspection equipment, any point In the camera coordinate system The normalized expression is: ； (A3) The relationship between the camera coordinate system and the captured image coordinate system is calculated as follows: , ,in, is a constant coefficient; (A4) When the world coordinate system and The relationship between the image coordinates and the camera angle when the axis is 0 is expressed as: The relationship between the image coordinates and the camera angle when the axis is 0 is expressed as: , , When the world coordinate system and The relationship between the image coordinates and the camera angle when the axis is 0 is expressed as: ； (A5) The camera's pitch angle, yaw angle, and rotation angle are related to The function of ; The step (S3) is specifically divided into the following steps: (B1) The pitch angle, roll angle and yaw angle of the aerial mobile inspection equipment are defined as , , , the coordinate system is established in the following way: the positive direction of the pitch axis points to the right side of the aerial mobile inspection device, the positive direction of the roll axis points to the front of the aerial mobile inspection device, and the positive direction of the yaw axis is perpendicular to the bottom of the aerial mobile inspection device; (B2) Calculate the total lift of the aerial mobile inspection equipment F It is expressed as: , in, Indicates the number of motors in the aerial mobile inspection equipment. is a constant coefficient affected by hardware materials, is the air density, is the motor area, Respectively n The propeller radius and rotational acceleration of each motor; (B3) The corresponding lift and , , The calculation formulas are: ， ， ， in, Indicates the distance between the motor and the center of gravity of the aerial mobile inspection equipment; (B4) Based on (A3), the pitch angle, roll angle and yaw angle of the aerial mobile inspection equipment are calculated as functions of lift: ; (B5) Calculate the distance between the smart sensor and the aerial mobile inspection equipment , , , are the horizontal distance and the vertical distance respectively, where The straight-line distance between the sensor and the obstacle. is the azimuth between the obstacle and the aerial mobile inspection equipment; (B6) Calculate the success rate of aerial mobile inspection equipment in identifying a target as follows: ； The step (S5) is specifically divided into the following steps: (D1) Define the ground mobile inspection equipment agent The state space and action space are: , ; Aerial mobile inspection equipment intelligent agent The state space and action space are: , ; The state space and action space of the ground fixed inspection equipment meta-agent are: , ; (D2) Define the ground mobile inspection agent In Status Next action The reward function is: , in, is a constant factor, the reward maximization is expressed as: , in, is the total number of ground mobile inspection agents, is the discount factor; (D3) Ground Mobile Inspection Agent In state The state value function of the strategy under is: ， in, For the agent in state action The policy function, is the strategy parameter, based on the above ground mobile inspection agent Choose the best strategy; (D4) Define the Aerial Mobile Inspection Equipment Agent In Status Next action The reward function is: ， in Angle Regarding the three lift functions, The probability of identifying a target for aerial mobile inspection equipment; Reward maximization is expressed as: , in is the total number of aerial mobile inspection agents, is the discount factor; (D5) Aerial Mobile Inspection Agent In state The state value function of the strategy under is: ， in For the agent in state action The policy function, is the strategy parameter. The above formula is used to evaluate the performance of the action under the state. Therefore, the aerial mobile inspection equipment agent Choosing the best strategy , the corresponding optimal value function is: ； (D6) Assume that the observation results of all local agents of ground mobile inspection and aerial mobile inspection are The observations are sent to the ground fixed inspection equipment. The meta-agent continuously updates the parameters through meta-learning. Get the optimal strategy; (D7) Strategies for ground-based fixed inspection equipment The advantage function is: ， in , , is the state transition probability; (D8) The loss function is defined as: ， in ; (D9) The policy gradient of the loss function is: ， in , ; (D10) The parameter update strategy for ground fixed inspection equipment is: ， is the meta-learning rate; (D11) The parameters of the ground fixed inspection equipment are further transmitted to the local agents of the ground mobile inspection equipment and the aerial mobile inspection equipment so that they can update their own optimal strategies.

2. The air-ground-ground integrated three-dimensional inspection method based on multi-intelligent voxel reinforcement learning according to claim 1 is characterized in that: Step (S4) is specifically divided into the following steps: (C1) Assume that the distance between any two of the aerial mobile inspection equipment, the ground mobile inspection equipment, and the ground fixed inspection equipment is , the azimuth is , and its communication includes line-of-sight transmission and non-line-of-sight transmission, the probability of the two types of communication is: ， in represents the probability of line-of-sight transmission, represents the non-line-of-sight transmission probability, is the signal steering vector, is a constant affected by the environment; (C2) Calculate the communication channel gain as: ， In the above formula, is the discount factor, is the speed of light, is the communication frequency, is the channel fading factor; (C3) The formula for calculating the communication rate between any two is: ， in, ， ， ， In the above formula, , is the communication rate between the aerial mobile inspection equipment and the ground mobile inspection equipment, is the communication rate between the aerial mobile inspection equipment and the ground fixed inspection equipment. is the communication rate between the ground mobile inspection equipment and the ground fixed inspection equipment; A is the signal transmission power, For noise, for The inverse of a function, is the block error probability caused by short data packet transmission, V is the channel divergence function, SNR is the signal-to-noise ratio, M is the block length, B is the channel width, t is the time coefficient, For each time slot duration; (C4) Communication delay is: ; B is the channel bandwidth, The number of pixel bits used to transmit video or pictures. is the video or image resolution, The number of devices communicating.

Images