Unmanned aerial vehicle automatic intelligent inspection method and inspection system

Abstract

The invention relates to the technical field of unmanned aerial vehicle inspection systems, and provides an unmanned aerial vehicle automatic intelligent inspection method and system, and the system comprises an unmanned aerial vehicle terminal, a ground management platform, a data transmission terminal, and an inspection regulation and control terminal. The ground management platform is used for planning an inspection task and a flight path for the unmanned aerial vehicle terminal; the unmanned aerial vehicle terminal is used for performing routing inspection according to the routing inspection task and the flight path and collecting sensing data required by the routing inspection task; the inspection regulation and control terminal is used for monitoring the state and environment change of the unmanned aerial vehicle terminal in real time and regulating and controlling the flight position of the unmanned aerial vehicle terminal in the process of collecting the sensing data by combining the sensing data; and the data transmission terminal is used for realizing sensing data transmission between the unmanned aerial vehicle terminal and the ground management platform based on the 5G wireless communication technology. The method has the effect of improving the unmanned aerial vehicle inspection data processing quality and efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle inspection systems, and in particular to an automatic intelligent unmanned aerial vehicle inspection method and an inspection system. Background Art

[0002] As a key industry supporting the quality of daily electricity consumption, the operation of hydropower plants directly impacts people's life satisfaction and well-being. To improve the quality of inspections of power supply equipment, drones should be flexibly employed for intelligent inspections, leveraging the advantages of these technologies to minimize labor costs, improve work efficiency, and further optimize power service quality by reducing workload.

[0003] Currently, inspections of hydropower plants and other critical infrastructure generally rely on traditional manual or semi-automated methods, which suffer from low efficiency, large coverage blind spots, incomplete data collection, and insufficient real-time response capabilities. In recent years, with the rapid development of drone, sensor, and data communication technologies, intelligent drone inspection systems have become an important approach to addressing these issues. While existing technologies offer data collection and inspection platforms based on multi-sensor fusion, they lack innovative calculation methods for unconventional object parameters, including dynamic environmental parameter assessment and multi-source data fusion. This limits the intelligence level of inspection systems and the accuracy of risk prediction.

[0004] Numerous drone inspection systems have been developed. These systems typically consist of a ground control terminal, an analysis terminal, and a drone terminal. The ground control terminal controls the drone terminal for inspections; the drone terminal collects inspection data and transmits it to the analysis terminal; and the analysis terminal analyzes the inspection data and synchronizes the analysis results and inspection data with the ground control terminal. Due to the relatively simple inspection and data processing processes of these drone inspection systems, and the lack of inspection control and data processing optimization, inspection data errors are prone to occur, resulting in reduced quality and efficiency of drone inspection data processing. Summary of the Invention

[0005] The purpose of the present invention is to address the shortcomings of the above-mentioned drone inspection system and propose an automatic intelligent drone inspection method and inspection system.

[0006] The present invention adopts the following technical solutions:

[0007] The automatic intelligent inspection system for drones includes a drone terminal, a ground management platform, a data transmission terminal, and an inspection and control terminal. The ground management platform is used to plan inspection tasks and flight trajectories for the drone terminal. The drone terminal is used to conduct inspections and collect sensor data required for the inspection tasks based on the inspection tasks and flight trajectories. The inspection and control terminal is used to monitor the status and environmental changes of the drone terminal in real time and, based on the sensor data, control the flight position of the drone terminal during the sensor data collection process. The data transmission terminal is used to realize sensor data transmission between the drone terminal and the ground management platform based on 5G wireless communication technology.

[0008] The data transmission terminal includes a raw data preprocessing module and a data transmission module; the raw data preprocessing module is used to preprocess the sensor data and generate target data; the data transmission module is used to transmit the target data to the ground management platform.

[0009] Optionally, the ground management platform includes an inspection task planning module, a flight trajectory planning module and a task monitoring module; the inspection task planning module is used to plan inspection tasks for the UAV terminal and determine the preset areas and target equipment that need to be inspected; the preset areas are flight areas pre-set by the administrator in the places specified by the inspection tasks; the flight trajectory planning module is used to plan the flight path and flight altitude for the UAV terminal; the task monitoring module is used to monitor the flight status of the UAV terminal and the execution of inspection tasks in real time.

[0010] Optionally, the UAV terminal includes a UAV body, a flight control module, a sensor data acquisition module and a task execution module; the flight control module, the sensor data acquisition module and the task execution module are installed on the UAV body; the flight control module is used to adjust the flight position and flight status of the UAV body according to the inspection tasks, flight path and flight altitude of the ground management platform; the task execution module is used to receive the inspection tasks issued by the ground management platform and determine the type of data required to be collected by the inspection tasks; the sensor data acquisition module is used to collect the required sensor data during the inspection process.

[0011] Optionally, the inspection and control terminal includes a real-time status monitoring module, an environmental change perception module and a flight position control module; the real-time status monitoring module is used to monitor the flight status of the UAV body in real time; the environmental change perception module is used to obtain changes in environmental data of the UAV body during flight; the flight position control module is used to adjust the position of the UAV body in a preset area according to changes in sensor data and environmental data.

[0012] Optionally, the flight position control module includes a feedback control index calculation submodule and a flight position control submodule within a preset area; the feedback control index calculation submodule is used to calculate the feedback control index of the drone body currently in the preset area according to changes in sensor data and environmental data; the flight position control submodule within the preset area is used to control the flight position of the drone body within the preset area according to the situation of the preset area, the flight state of the drone body and the feedback control index;

[0013] When the feedback control index calculation submodule calculates, the following formula is satisfied:

[0014]

[0015] Among them, G represents the feedback control index of the drone body currently in the preset area; K represents the shrinkage factor; X i represents the raw data value of the i-th sensor in the sensor data acquisition module; N represents the total number of sensor types in the sensor data acquisition module; α i represents the amplification factor of the i-th sensor; Y i represents the noise influence coefficient of the i-th sensor in the sensor data acquisition module. The higher the noise level in the raw data of the i-th sensor, the greater the corresponding noise influence coefficient; Z i Represents the reliability score of the i-th sensor in the sensor data acquisition module; Max i Represents the preset maximum value of the sum of the i-th sensor, which is used to standardize the value of the sensor sum; β i represents the nonlinear control coefficient of the i-th sensor; γ i Indicates the adjustment coefficient of the i-th sensor; Q i represents the sensitivity coefficient of the i-th sensor; w represents the weight coefficient of the product term;

[0016] When G<g ref When the flight position control submodule in the preset area controls the flight position of the drone body in the preset area according to the situation of the preset area and the flight status of the drone body, the G value is optimized until G≥g ref ;g ref Indicates the evaluation threshold of the drone's flight position when the drone terminal collects sensor data.

[0017] Optionally, the raw data preprocessing module includes a multi-type sensor data preprocessing submodule and a target data generation submodule; the multi-type sensor data preprocessing submodule is used to preprocess the raw data of various sensors; the target data generation submodule is used to package the preprocessed data into target data.

[0018] Optionally, the multi-type sensor data preprocessing submodule includes an image data preprocessing unit, a temperature data preprocessing unit, a vibration data preprocessing unit and a humidity data preprocessing unit; the image data preprocessing unit is used to perform gradient enhancement processing and noise reduction processing on the original data of the image sensor, and generate an image enhancement index; the temperature data preprocessing unit is used to perform deviation degree analysis on the original data of the temperature sensor, and generate a temperature global index; the vibration data preprocessing unit is used to amplify and suppress interference on the original data of the vibration sensor, and generate a vibration global index; the humidity data preprocessing unit is used to perform data smoothing and screening on the original data of the humidity sensor, and generate a humidity global index; the target data generation submodule is used to package each type of preprocessed sensor data, image enhancement index, temperature global index, vibration global index and humidity global index into target data.

[0019] The automatic intelligent inspection method of a drone is applied to the automatic intelligent inspection system of a drone as described above, and the automatic intelligent inspection method of a drone includes:

[0020] S1, planning inspection tasks and flight trajectories;

[0021] S2, based on the inspection task and flight trajectory, conducts inspections and collects sensor data required by the inspection task;

[0022] S3, real-time monitoring of the status of the UAV terminal and environmental changes, and combined with sensor data to control the flight position of the UAV terminal during the process of collecting sensor data;

[0023] S4, based on 5G wireless communication technology, realizes sensor data transmission between drone terminals and ground management platforms.

[0024] The beneficial effects achieved by the present invention are:

[0025] 1. Through the setting of drone terminals, ground management platforms, data transmission terminals and inspection and control terminals, the system can realize the planning, execution and real-time monitoring of inspection tasks, which is conducive to improving the autonomy and data collection efficiency of drones during the inspection process, and then realize efficient data transmission through 5G wireless communication technology, which is conducive to the realization of intelligent drones, remote real-time control inspection tasks, and improve the quality and efficiency of drone inspection data processing.

[0026] 2. Through the setting of the inspection task planning module, flight trajectory planning module and task monitoring module in the ground management platform, it is possible to accurately plan inspection tasks, flight paths and altitudes, and monitor the flight status of drones in real time, which is conducive to ensuring the accurate execution of inspection tasks and the optimization of flight paths, thereby improving the safety and efficiency of drone task execution, thereby helping to reduce errors and improve the accuracy of inspections.

[0027] 3. Through the settings of the flight control module, sensor data acquisition module and task execution module in the UAV terminal, it can autonomously adjust the flight position and status under the guidance of the inspection task, and collect the required sensor data in real time, which is conducive to the UAV to flexibly adjust the working status according to the task requirements, thereby ensuring the accuracy and stability of task execution, which is conducive to improving the inspection quality and reducing human intervention.

[0028] 4. Through the setting of the real-time status monitoring module, environmental change perception module and flight position control module in the inspection and control terminal, the drone status and environmental changes can be monitored in real time, and the position of the drone can be adjusted according to the data, which is conducive to flexibly adjusting the flight position according to the actual flight environment changes, thereby enhancing the adaptability and emergency response capabilities of the inspection, which is conducive to improving the adaptability and flexibility of the drone in complex environments.

[0029] 5. Through the setting of multi-type sensor data preprocessing submodule and target data generation submodule in the raw data preprocessing module, corresponding preprocessing and target data generation can be performed for data from different sensors, thereby improving the quality and consistency of various types of data, which is conducive to reducing inconsistencies between data, and then ensuring the accuracy and uniformity of data, which is conducive to efficient integration and further processing of data, and improving the quality and efficiency of drone inspection data processing.

[0030] 6. Through the setting of the image data preprocessing unit, temperature data preprocessing unit, vibration data preprocessing unit and humidity data preprocessing unit in the multi-type sensor data preprocessing sub-module, specific preprocessing can be performed on the data of each sensor, thereby improving the quality of each type of sensor data, which is conducive to enhancing the accuracy and reliability of the data, and thus improving the overall intelligent inspection capability of the system, thereby helping to enhance the multi-functional application effect of drones in various complex environments, and improving the quality and efficiency of drone inspection data processing.

[0031] To further understand the features and technical contents of the present invention, please refer to the following detailed description and drawings of the present invention. However, the drawings provided are only for reference and illustration and are not intended to limit the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a schematic diagram of the overall structure of the present invention;

[0033] Figure 2 Schematic diagram of the structure of the flight position control module in the present invention;

[0034] Figure 3 This is a schematic diagram of the effect of running part of the program code for calculating the G value in the present invention;

[0035] Figure 4 This is a schematic diagram showing the effect of running the supplementary program code for the G value calculation program code of the present invention;

[0036] Figure 5 Schematic diagram of the structure of the raw data preprocessing module in the present invention;

[0037] Figure 6 This is a flow chart of the method for the automatic intelligent inspection method of a drone in the present invention;

[0038] Figure 7 A schematic diagram summarizing the entire process of system operation in another embodiment of the present invention;

[0039] Figure 8 This is a schematic diagram of the on-site inspection effect of the drone body in another embodiment of the present invention. DETAILED DESCRIPTION

[0040] The following is an explanation of the embodiments of the present invention through specific embodiments. Those skilled in the art can understand the advantages and effects of the present invention from the contents disclosed in this specification. The present invention can be implemented or applied through other different specific embodiments, and the details in this specification can also be modified and changed based on different viewpoints and applications without departing from the spirit of the present invention. In addition, the drawings of the present invention are only for simple schematic illustrations and are not depicted according to actual dimensions. It is stated in advance. The following embodiments will further explain the relevant technical contents of the present invention in detail, but the disclosed contents are not intended to limit the scope of protection of the present invention.

[0041] Embodiment 1: This embodiment provides an automatic intelligent inspection system for drones. The corresponding practical application scenarios in this embodiment are as follows: the drone body in the system performs inspection tasks and flies to a corresponding place with a large enough space, such as a hydropower plant. The space in the place is pre-divided into several preset areas by the administrator according to the equipment on site. The preset areas belong to large areas, which at least allow the drone body to adjust its position without obstacles within the preset areas. The drone body needs to complete inspection actions in each preset area, collect sensor data, pre-process the sensor data and complete the transmission. Combined with Figure 1 As shown, the automatic intelligent inspection system for drones includes a drone terminal, a ground management platform, a data transmission terminal, and an inspection and control terminal; the ground management platform is used to plan inspection tasks and flight trajectories for the drone terminal; the drone terminal is used to perform inspections and collect sensor data required for the inspection tasks based on the inspection tasks and flight trajectories; the inspection and control terminal is used to monitor the status and environmental changes of the drone terminal in real time, and control the flight position of the drone terminal during the sensor data collection process based on the sensor data; the data transmission terminal is used to realize sensor data transmission between the drone terminal and the ground management platform based on 5G wireless communication technology;

[0042] The data transmission terminal includes a raw data preprocessing module and a data transmission module; the raw data preprocessing module is used to preprocess the sensor data and generate target data; the data transmission module is used to transmit the target data to the ground management platform.

[0043] Optionally, the ground management platform includes an inspection task planning module, a flight trajectory planning module and a task monitoring module; the inspection task planning module is used to plan inspection tasks for the UAV terminal and determine the preset areas and target equipment that need to be inspected; the preset areas are flight areas pre-set by the administrator in the places specified by the inspection tasks; the flight trajectory planning module is used to plan the flight path and flight altitude for the UAV terminal; the task monitoring module is used to monitor the flight status of the UAV terminal and the execution of inspection tasks in real time.

[0044] Optionally, the UAV terminal includes a UAV body, a flight control module, a sensor data acquisition module and a task execution module; the flight control module, the sensor data acquisition module and the task execution module are installed on the UAV body; the flight control module is used to adjust the flight position and flight status of the UAV body according to the inspection tasks, flight path and flight altitude of the ground management platform; the task execution module is used to receive the inspection tasks issued by the ground management platform and determine the type of data required to be collected by the inspection tasks; the sensor data acquisition module is used to collect the required sensor data during the inspection process.

[0045] Optionally, the inspection and control terminal includes a real-time status monitoring module, an environmental change perception module and a flight position control module; the real-time status monitoring module is used to monitor the flight status of the UAV body in real time; the environmental change perception module is used to obtain changes in environmental data of the UAV body during flight; the flight position control module is used to adjust the position of the UAV body in a preset area according to changes in sensor data and environmental data.

[0046] Optional, combined Figure 2 As shown, the flight position control module includes a feedback control index calculation submodule and a flight position control submodule within a preset area; the feedback control index calculation submodule is used to calculate the feedback control index of the UAV body currently in the preset area according to changes in sensor data and environmental data; the flight position control submodule within the preset area is used to control the flight position of the UAV body within the preset area according to the situation of the preset area, the flight state of the UAV body and the feedback control index;

[0047] When the feedback control index calculation submodule calculates, the following formula is satisfied:

[0048]

[0049] Among them, G represents the feedback control index of the drone body currently in the preset area; K represents the reduction factor, which is set by the administrator based on experience. In this embodiment, the value selection rule of this value is as follows: the more types of sensors there are, the larger the K value is, K = 1*10 M , M = number of sensor types - 1; X i represents the raw data value of the i-th sensor in the sensor data acquisition module; in this embodiment, one sensor type corresponds to one sensor; N represents the total number of sensor types in the sensor data acquisition module; α i represents the amplification factor of the i-th sensor. In this embodiment, the selection rule of this value is as follows: the more the number of failures of the i-th sensor is, the smaller the amplification factor is. i =1-failure times / 10. The maximum number of sensor failures is 9. When the sensor fails for the 10th time, it needs to be replaced.

[0050] Y i It represents the noise influence coefficient of the i-th sensor in the sensor data acquisition module. The higher the noise level in the raw data of the i-th sensor, the larger the corresponding noise influence coefficient:

[0051]

[0052] Among them, λ i represents the noise impact factor of the i-th sensor; σ ei μ represents the standard deviation of the measurement error of the i-th sensor, which can be obtained by experimental verification and calculation: in a laboratory environment, a limited number of repeated measurements of the sensor are performed, such as calibration point detection, to obtain the distribution of measurement errors and calculate the standard deviation. It can also be obtained from the technical documentation provided by the manufacturer; i represents the average measurement value of the i-th sensor, that is, the average value of the historical sensor data obtained by the sensor in the corresponding preset area; σ envi σ represents the standard deviation of the environmental noise of the i-th sensor. It mainly measures the interference of environmental factors such as temperature, humidity, vibration, and light on the sensor measurement results. For example, the additional error of the temperature sensor under different external temperatures, the fluctuation of the humidity sensor under different humidity environments, etc. It is obtained by testing the noise level of the sensor under different environmental conditions, and then calculating its standard deviation by repeatedly measuring the sensor output under different environmental conditions a limited number of times. instiη1 represents the standard deviation of the intrinsic noise of the i-th sensor. It mainly measures the inherent noise of the sensor hardware during operation, such as thermal noise of electronic components and interference caused by loose mechanical structure. It is usually related to the internal design of the sensor, chip selection, and circuit stability. It is obtained by testing the stability of the sensor under long-term operation or different operating conditions. It can be obtained by long-term monitoring in a stable environment with constant sensor input, recording the fluctuation of the sensor output, and then calculating it using the standard deviation. η1, η2, and η3 represent different weight coefficients. The specific value determination method can be, but is not limited to: 1. Setting based on the administrator's experience: If the administrator knows which of the measurement error, environmental noise, and instrument noise is the main conflicting item or the dominant item, then a larger weight can be assigned accordingly. For example, a larger weight can be assigned to a sensor that is easily affected by environmental interference. 2. Data regression after a limited number of experiments: Retrieve the sensor's historical measurement data under different environments and use regression methods, such as the least squares method, to fit a set of optimal weights based on the actual measured errors. 3. Iterative optimization: During actual use, continuously record the sensor output and the actual value and iteratively adjust each weight coefficient until the performance requirements set by the administrator are met.

[0053] Z i Represents the reliability score of the i-th sensor in the sensor data acquisition module:

[0054]

[0055] Among them, δ i Represents the reliability influencing factor of the i-th sensor; Lifespan i Indicates the service life of the i-th sensor, which is specifically the marked working years in this embodiment and is obtained from the technical documents provided by the manufacturer; Max-Lifespan i represents the maximum service life of the i-th sensor, obtained from the technical documentation provided by the manufacturer; μ i represents the average measurement value of the i-th sensor, that is, the average value of the historical sensor data obtained by the sensor in the corresponding preset area; σ cali It represents the standard deviation of the calibration error of the i-th sensor, which measures the error fluctuation between the measured value and the standard reference value during the sensor calibration process. It is obtained by comparing the sensor with a known standard reference source under controlled conditions. It represents the inherent measurement error of the sensor and is usually related to the hardware accuracy of the sensor itself and the accuracy of the calibration process. The specific calculation method is: the difference between multiple measurements and the known standard value is calculated, and then its standard deviation is calculated. P represents the number of calibrations, X j,cal represents the measured value of the standard reference source for the jth measurement, X refIndicates the standard value of the standard reference source; σ envi The standard deviation of the environmental noise of the i-th sensor is mainly used to measure the interference of environmental factors such as temperature, humidity, vibration, and light on the sensor measurement results. For example, the additional error of the temperature sensor under different external temperatures, the fluctuation of the humidity sensor under different humidity environments, etc. It is obtained by testing the noise level of the sensor under different environmental conditions, and then calculating its standard deviation by repeatedly measuring the sensor output under different environmental conditions a limited number of times. i The maximum stability score of the i-th sensor indicates how long and how wide a range of changes the sensor can remain stable. It is obtained from the technical documentation provided by the manufacturer or through long-term stability testing: X t,max Represents the measured value of the sensor at time t, t is generally greater than 1 year, X t0 Represents the measurement value of the sensor in the initial stable state; φ1, φ2 and φ3 represent different proportional coefficients respectively. The specific method of confirming the value can be, but is not limited to: 1. Setting according to the administrator's experience: If the administrator knows which of the items in the above formula is the main contradiction or the dominant item, then a larger proportion is set accordingly; 2. Data regression after a limited number of experiments: retrieve the historical measurement data of the sensor in different environments, and use regression methods such as the least squares method to fit a set of optimal weight groups based on the errors obtained from actual measurements; 3. Iterative optimization: During actual use, the sensor output and true value are continuously recorded, and each weight coefficient is iteratively corrected until the performance requirements set by the administrator are met.

[0056] Max i It represents the preset maximum value of the sum of the i-th sensor, which is used to standardize the value of the sensor sum. The specific value is set by the administrator based on experience. The selection rule is: the larger the average value of the historical sum of the i-th sensor, the higher the Max. i The larger the value, the higher the Max value. i The digits of the value are consistent with the digits of the average value of the historical summation item of the i-th sensor; β i It represents the nonlinear control coefficient of the i-th sensor, which is used to describe the degree of signal amplification caused by the nonlinear effect during the measurement process of the sensor. It represents the nonlinear amplification effect of the sensor under different working conditions. γ i It represents the adjustment coefficient of the i-th sensor, which is used to describe the inhibitory effect of environmental conditions on its output signal during sensor measurement. It reflects the nonlinear suppression capability of the sensor under different environmental conditions. Q i represents the sensitivity coefficient of the i-th sensor, w represents the weight coefficient of the product term, and there are three alternative values: 0.2, 0.25 and 0.3. When the number of sensor types is less than 4, 0.3 is selected; when the number of sensor types is greater than or equal to 4, 0.25 is selected; when the number of sensor types is greater than or equal to 4 and less than 7, 0.25 is selected; when the number of sensor types is greater than or equal to 7, 0.2 is selected.

[0057] When G<g ref When the flight position control submodule in the preset area controls the flight position of the drone body in the preset area according to the situation of the preset area and the flight status of the drone body, the G value is optimized until G≥g ref ;g ref It represents the judgment threshold of the drone’s flight position when the drone terminal collects sensor data. It is set by the administrator based on experience. The g of each preset area ref All are preset by the administrator in advance. The preset method can be, but is not limited to: the administrator holds the data collection module and moves to the location with the best data quality within the corresponding preset area to determine the location. The strategies for adjusting the flight position of the drone body within the same preset area can include, but are not limited to, the following directions:

[0058] 1. Adjust the flight altitude first: The sensor's data collection effect may vary with the flight altitude. You can adjust the altitude of the drone to find different flight altitudes within the same preset area to improve data quality. It is important to note that the flight altitude adjustment process is adjusted within the effective altitude range. That is, during the adjustment process, the data collection of various types of sensors remains normal. Sensors that were collecting data normally before the altitude adjustment will continue to collect data normally throughout the adjustment process.

[0059] Adjustment method: Adjust the flight altitude, such as raising or lowering it, and recalculate the G value until it reaches the standard.

[0060] 2. Adjust the flight direction: Within the preset area, the sensor data quality may degrade due to changes in viewing angle, lighting, or wind speed. By adjusting the flight direction, you can change the sensor's collection angle, thereby improving data quality. It should be noted that the flight direction adjustment process is adjusted within the effective altitude range, that is, during the adjustment of the flight direction, the collection work of various types of sensors remains normal. Sensors that were collecting data normally before the flight direction adjustment continue to collect data normally throughout the entire process of adjusting the flight direction.

[0061] Adjustment method: Change the flight angle or direction of the drone and recalculate the G value until it reaches the standard.

[0062] 3. Finally, fine-tune the flight position: Within the preset area, the drone can find the most suitable collection point by fine-tuning its flight position;

[0063] Adjustment method: Based on the feedback from the sensor data, make small lateral or longitudinal movements and recalculate the G value until the standard is met.

[0064] It should be noted that when calculating the G value, the value of the number of sensor types N is confirmed by the number of sensors that are working normally, that is, the sensor data value substituted into the calculation is not zero.

[0065] The following is an example of the G-value calculation process, which illustrates and supplements the formula algorithm for the feedback control index of the drone body currently in the preset area:

[0066] The known conditions are: g ref =0.025, K=1000, w=0.25, N=4, respectively including temperature sensor, vibration sensor, image sensor and humidity sensor;

[0067] Temperature sensor: measurement value X1 = 25°C, maximum preset value of summation item Max1 = 10 11 , noise impact factor λ1=0.006, reliability impact factor δ1=0.1; measurement error standard deviation σ e1 =0.2℃, standard deviation of environmental noise σ env1 =0.1℃, instrument noise standard deviation σ inst1 =0.05℃;

[0068] Vibration sensor: measurement value X2 = 0.1m / s 2 , the maximum preset value of the summation term Max2 = 10, the noise impact factor λ2 = 0.0115, the reliability impact factor δ2 = 0.3; the measurement error standard deviation σ e2 =0.02m / s 2 , standard deviation of environmental noise σ env2 =0.01m / s 2 , instrument noise standard deviation σ inst2 =0.005m / s 2 ;

[0069] Image sensor: measurement value X3 = 0.8, measurement value is standardized image data, that is, standardized image pixels, maximum preset value of sum item Max3 = 1, noise impact factor λ3 = 0.07, reliability impact factor δ3 = 0.2; measurement error standard deviation σ e3 =0.1, standard deviation of environmental noise σ env3 =0.05, instrument noise standard deviation σ inst3 =0.02;

[0070] Humidity sensor: measured value X4 = 45%, summation maximum preset value Max4 = 1018 , noise impact factor λ4=0.000333, reliability impact factor δ4=0.15; measurement error standard deviation σ e4 =0.03%, standard deviation of environmental noise σ env4 =0.01%, instrument noise standard deviation σ inst4 =0.005%.

[0071] Substituting into the formula we get:

[0072]

[0073] Z1≈0.9091, Z2≈0.7692, Z3≈0.8333, Z4≈0.8696;

[0074] For the convenience of calculation and to show the use process of the calculation formula, the following β i , γ i , Q i Adjustments have been made, and in actual application, calculations must be performed according to the above formulas and explanations;

[0075] Due to the characteristics of the sin function, when the input value causes the result of the sin function to be negative, the input value needs to be adjusted first. For example, when the measured value of the temperature sensor is 25℃, due to β i , γ i , Q i The adjusted value has been fixed. The specific adjustment method is: adjustment value = 25*0.01 = 0.25; in actual application, it can be combined with β i The value of is adjusted so that sin(β i X i ) is a positive number.

[0076] Temperature sensor product term:

[0077] Vibration sensor product term:

[0078] Image sensor product term:

[0079] Humidity sensor product term:

[0080] Total value of product term = (1.2475 × 1.0749 × 1.402 × 1.707) 0.25 ≈1.214;

[0081]

[0082] Temperature sensor summation term:

[0083] Vibration sensor summation term:

[0084] Image sensor summation term:

[0085] Humidity sensor summation term:

[0086] Total value of the sum = 0.321 + 0.00907 + 0.8215 + 15.9 = 17.0516;

[0087] G <g ref The flight position control submodule in the preset area controls the flight position of the UAV body in the preset area according to the above method, and optimizes the G value until G≥g ref .

[0088] Combine Figure 3 and Figure 4 The following is the core program code that implements the above calculation example. This program code is helpful for quickly calculating the G value and is also easy to modify the code to adapt to various calculation scenarios:

[0089] import numpy as np

[0090] #Known data of the sensor

[0091] #Temperature Sensor

[0092] mu_1=25# temperature sensor measurement value (℃)

[0093] lambda_1=0.006#Temperature sensor noise impact factor

[0094] beta_1=0.1#Temperature sensor reliability impact factor

[0095] max_1=10**11#Preset maximum value of the temperature sensor summation item

[0096] #Vibration Sensor

[0097] mu_2=0.1#Vibration sensor measurement value (m / s^2)

[0098] lambda_2=0.0115#Vibration sensor noise impact factor

[0099] beta_2=0.3#Vibration sensor reliability impact factor

[0100] max_2=10#Preset maximum value of the vibration sensor summation item

[0101] #Image Sensor

[0102] mu_3=0.8#Image sensor measurement value (normalized value)

[0103] lambda_3=0.070#Image sensor noise impact factor

[0104] beta_3=0.2#Image sensor reliability impact factor

[0105] max_3=1#Preset maximum value of the image sensor summation item

[0106] #Humidity Sensor

[0107] mu_4=45#Humidity sensor measurement value (%)

[0108] lambda_4=0.000333#Humidity sensor noise impact factor

[0109] beta_4=0.15#Humidity sensor reliability impact factor

[0110] max_4=10**18#Preset maximum value of the humidity sensor summation item

[0111] #Sensor measurement value

[0112] measurements=[mu_1,mu_2,mu_3,mu_4]

[0113] lambdas=[lambda_1,lambda_2,lambda_3,lambda_4]

[0114] betas=[beta_1,beta_2,beta_3,beta_4]

[0115] max_values=[max_1,max_2,max_3,max_4]

[0116] #Define the calculation formula for the sum

[0117] def compute_Yi(lambda_i):

[0118] """Calculate the noise impact coefficient Yi"""

[0119] return 1 / (1+lambda_i)

[0120] def compute_Zi(beta_i):

[0121] """Calculate the reliability influence coefficient Zi"""

[0122] return 1 / (1+beta_i)

[0123] def compute_sum_term(X_i,Y_i,Z_i):

[0124] """Calculate the sum of each sensor"""

[0125] return(np.exp(X_i)-np.exp(-X_i)) / (np.arctan(Y_i)+np.sqrt(Z_i+1))defcompute_product_term(W_i,beta_i,gamma_i,Q_i):

[0126] """Calculate the product term for each sensor"""

[0127] return 1+(np.sin(beta_j*X_i)) / (np.cosh(gamma_i*Q_i))

[0128] # Calculate Yi and Zi for each sensor Yi = [compute_Yi(lambda_i)forlambda_iinlambdas]

[0129] Zi=[compute_Zi(beta_i)for beta_iin betas]

[0130] #Calculate the sum

[0131] sum_terms = []

[0132] fori,measurementin enumerate(measurements):

[0133] sum_term=compute_sum_term(measurement,Yi[i],Zi[i])

[0134] sum_terms.append(sum_term)

[0135] # Calculate the normalized sum

[0136] normalized_sum_terms=[sum_term / max_values[i]fori,sum_terminumerate(sum_terms)]

[0137] #Calculate the total value of the sum

[0138] total_sum_term=np.sum(normalized_sum_terms)

[0139] #Calculate the product term (assuming the weight coefficient is 0.25, gamma_i=0.8, Q_i=0.02)

[0140] product_terms=[]

[0141] gamma_i=0.8#Assuming gamma_i value

[0142] Q_i=0.02#Assuming Q_i value

[0143] foriin range(4):

[0144] product_term=compute_product_term(measurements[i], betas[i], gamma_i, Q_i)

[0145] product_terms.append(product_term)

[0146] #Assume that the product term weight of each sensor is 0.25

[0147] w_i=0.25

[0148] adjusted_product_term=np.prod(np.array(product_terms)**w_i)

[0149] #scaling factor

[0150] K=1000

[0151] #Calculate G value

[0152] G=(1 / K)*total_sum_term*adjusted_product_term

[0153] # Output the final result

[0154] print(f"The final calculated G value: {G}").

[0155] In summary, the data transmission terminal includes a raw data preprocessing module and a data transmission module, which can ensure that the collected sensor data is transmitted to the ground management platform after preprocessing, ensure the accuracy of the data, and thus improve the efficiency of data transmission. Real-time status monitoring and adjustment: The inspection and control terminal combines the real-time monitoring module and the environmental change perception module to adjust the flight position of the drone in time according to changes in environmental data, optimize the data collection effect, and thus reduce the inspection error caused by environmental factors. Feedback control index: Through the setting of the feedback control index calculation submodule and the flight position control module, the flight position can be adjusted according to sensor data and environmental changes, and the flight position of the drone in the preset area can be optimized through algorithms to ensure the accuracy and stability of the data collection process, reduce unnecessary flight paths and position adjustments, and improve the quality and efficiency of drone inspection data processing.

[0156] Optional, combined Figure 5 As shown, the raw data preprocessing module includes a multi-type sensor data preprocessing submodule and a target data generation submodule; the multi-type sensor data preprocessing submodule is used to preprocess the raw data of various sensors; the target data generation submodule is used to package the preprocessed data into target data.

[0157] Optionally, the multi-type sensor data preprocessing submodule includes an image data preprocessing unit, a temperature data preprocessing unit, a vibration data preprocessing unit and a humidity data preprocessing unit; the image data preprocessing unit is used to perform gradient enhancement processing and noise reduction processing on the original data of the image sensor, and generate an image enhancement index; the temperature data preprocessing unit is used to perform deviation degree analysis on the original data of the temperature sensor, and generate a temperature global index; the vibration data preprocessing unit is used to amplify and suppress interference on the original data of the vibration sensor, and generate a vibration global index; the humidity data preprocessing unit is used to perform data smoothing and screening on the original data of the humidity sensor, and generate a humidity global index; the target data generation submodule is used to package each type of preprocessed sensor data, image enhancement index, temperature global index, vibration global index and humidity global index into target data.

[0158] The automatic intelligent inspection method of UAV is applied to the automatic intelligent inspection system of UAV as shown in the figure. Figure 6 As shown, the automatic intelligent inspection method of the drone includes:

[0159] S1, planning inspection tasks and flight trajectories;

[0160] S2, based on the inspection task and flight trajectory, conducts inspections and collects sensor data required by the inspection task;

[0161] S3, real-time monitoring of the status of the UAV terminal and environmental changes, and combined with sensor data to control the flight position of the UAV terminal during the process of collecting sensor data;

[0162] S4, based on 5G wireless communication technology, realizes sensor data transmission between drone terminals and ground management platforms.

[0163] Example 2: This example includes all the contents of Example 1 and provides an automatic intelligent inspection system for drones. When the drone body is in a flight position such that G≥g ref , the multi-type sensor data preprocessing submodule starts working; when the image data preprocessing unit is working, the raw data collected by the image sensor is first subjected to gradient enhancement processing and noise reduction processing, so that the raw image data is converted into optimized image data; the gradient enhancement processing and noise reduction processing methods can be, but are not limited to: input image; perform gradient calculation to obtain gradient amplitude; perform gradient enhancement according to the amplification factor and gradient amplitude; output image as noise image; input noise image; select mean filter for filtering; output filtered image as optimized image data. The process of generating image enhancement index satisfies the following formula:

[0164]

[0165] Among them, I * Represents the image enhancement index at the current moment; I * The larger the value, the more significant the enhancement effect; N represents the total number of pixels in the sampled image, and the image in the original data of the image sensor is sampled; ΔI i Represents the gradient change value of the i-th pixel in the image, obtained from the gradient enhancement process, G td represents the gradient amplitude, G x Represents the horizontal gradient of the i-th pixel, which is obtained by convolution of the pixel gray value and the Sobel operator. y It represents the vertical gradient of the ith pixel, which is obtained by convolution of the pixel gray value and the Sobel operator; ∈ represents a minimum constant, which is used to avoid instability caused by division by zero or square root during calculation, and its value range is 10 -3 to 10 -5 The specific value is set by the administrator based on experience. The empirical value selection rule is: when processing images with severe noise, ∈ can be appropriately increased to avoid excessive amplification of noise during gradient calculation; for images with good quality and less noise, ∈ can be set to a smaller value, typically 10 -3 : Used in lower noise images to reduce excessive amplification of noise when enhancing edges. The typical value is 10 -5 : It is used in scenes with clear images and low noise, which helps to improve detail accuracy; κi It represents the noise characteristic adjustment factor of the i-th pixel in the image. Common values are: 0.1 and 0.01. 0.1 is used for images with strong noise, which helps to reduce noise; 0.01 is used for images with less noise, which can improve the image details. In this embodiment, κ i =0.1. The target data includes optimized image data and image enhancement indicators and their calculation formulas, which helps reduce storage space and improve storage and transmission efficiency and accuracy. It also facilitates data restoration and acquisition of various data in the image preprocessing process. The existence of image enhancement indicators also makes it easier for administrators and systems to judge the quality of image data, thereby improving acquisition accuracy and efficiency, and further improving the accuracy and efficiency of the entire data processing process from data acquisition to transmission.

[0166] When the temperature data preprocessing unit is working, the original data collected by the temperature sensor is first analyzed for deviation. The deviation analysis method can be, but is not limited to: input the temperature data set; calculate the mean and standard deviation of the temperature data set; calculate the deviation degree of each data point from the mean: deviation degree = absolute value of the difference between the temperature data and the mean / standard deviation; set a deviation threshold and compare the deviation degree, and determine the temperature data with a deviation greater than the deviation threshold as abnormal data and filter it out. When generating the global temperature index, the following formula is satisfied:

[0167]

[0168] Among them, T * Represents the global temperature index at the current moment; the smaller the global temperature index, the higher the quality of the temperature data; T i represents the i-th temperature data in the temperature data set collected by the temperature sensor at the current moment; T ref It represents the standard reference operating temperature value of the inspection object; N represents the total number of temperature data in the temperature data set; the target data includes the filtered temperature data set and the global temperature index and its calculation formula, which is conducive to reducing storage space and improving the efficiency and accuracy of storage and transmission. The existence of the global temperature index also makes it easier for administrators and systems to judge the quality of temperature data, so as to improve the accuracy and efficiency of data collection, thereby improving the accuracy and efficiency of the entire data processing process from data collection to transmission.

[0169] When the vibration data preprocessing unit is working, the raw data collected by the vibration sensor is first amplified and interference suppressed. The amplification and interference suppression methods can be, but are not limited to: inputting a vibration data set; selecting an amplification factor for amplitude amplification; outputting the amplified vibration data as input for interference suppression; selecting a low-pass filter for filtering; and outputting the vibration data after interference suppression. The process of generating the global vibration index satisfies the following formula:

[0170]

[0171] Among them, V * Represents the global vibration index at the current moment; V * The larger the value, the stronger the amplitude and fluctuation of the vibration signal, which usually means that the vibration is more severe, which may indicate that there is greater mechanical stress, equipment failure or abnormal status in the system; i Indicates the amplification factor selected when performing amplitude amplification. In this embodiment, 1.5 is selected; V i Represents the amplitude of the i-th vibration data point in the vibration dataset. The target data includes the amplified and interference-suppressed vibration dataset, along with the global vibration index and its calculation formula. This helps reduce storage space and improves the efficiency and accuracy of storage and transmission. The global vibration index also facilitates administrators and systems in assessing the quality of vibration data, improving data collection accuracy and efficiency, and ultimately, the accuracy and efficiency of the entire data processing process, from data collection to transmission.

[0172] When the humidity data preprocessing unit is working, the raw data collected by the humidity sensor is first smoothed and filtered. The data smoothing and filtering method can be, but is not limited to: input a humidity data set; perform sliding window averaging on the humidity data with a window size of 3, that is, in the humidity data set, each data point is converted into the average of the current data point and the two data points before and after it; and output the smoothed humidity data set. When generating the global humidity index, the following formula is satisfied:

[0173]

[0174] Among them, H * Represents the global humidity index at the current moment; H * A larger value usually indicates a larger difference between the humidity data and the reference value, and a stronger data fluctuation, which may reflect a larger error or sensor instability; N represents the total number of data points in the humidity data set; H i Represents the humidity data of the i-th data point in the humidity dataset at the current moment; H ref Represents the standard reference operating humidity value of the inspection object; the target data includes a smoothed humidity data set and a global humidity index and its calculation formula, which is conducive to reducing storage space and improving the efficiency and accuracy of storage and transmission. The existence of the global humidity index also makes it easier for administrators and systems to judge the quality of humidity data, so as to improve the accuracy and efficiency of data collection, thereby improving the accuracy and efficiency of the entire data processing process from data collection to transmission.

[0175] It should be noted that when using the above indicators for judgment, the following methods may be used, but are not limited to: 1. Threshold comparison method, in which a threshold is preset and compared with the indicator to make a judgment; 2. Experience observation method, in which the administrator observes and makes judgments based on experience; 3. Historical mean comparison method, in which the current indicator value is compared with the historical mean of the indicator and a judgment is made based on the specific degree of deviation.

[0176] The working process of the system in this embodiment can be summarized as follows: Figure 7 and Figure 8 As shown, the drone body is first controlled to fly to one of the corresponding preset areas in the venue. The drone body performs the inspection task, and each sensor starts to work normally. At this time, the G value is continuously calculated and then the flight position of the drone body in the preset area is adjusted. When the G value meets the conditions, the next step of data collection and data processing begins. The various units in the multi-type sensor data preprocessing submodule start working, and then the corresponding target data is obtained after data processing, thereby completing accurate, efficient and energy-saving data transmission.

[0177] In summary, image data preprocessing: The image sensor's raw data is optimized through gradient enhancement and noise reduction, reducing the impact of noise on image quality and ensuring the accuracy of the image enhancement effect. Image enhancement metrics help the system assess image data quality and further improve image data acquisition accuracy. Temperature data deviation analysis: The deviation analysis algorithm accurately detects and filters outliers in temperature data, improving the reliability of temperature data and ensuring the system accurately collects valid temperature data during environmental monitoring, reducing the impact of errors and interference. Vibration data amplification and interference suppression: Vibration sensor data is amplified and subjected to interference suppression, improving the readability and accuracy of vibration signals. This helps the system accurately identify the health status of mechanical equipment and potential faults, providing early warnings. Humidity data smoothing and screening: Humidity data is processed through sliding window averaging to smooth fluctuations, reduce interference from environmental noise, and improve the accuracy and stability of humidity data. This ensures stable and reliable real-time environmental data acquisition during inspections, enhancing the quality and efficiency of drone inspection data processing.

[0178] The contents disclosed above are only preferred feasible embodiments of the present invention and do not limit the scope of protection of the present invention. Therefore, all equivalent technical changes made using the contents of the present invention description and drawings are included in the scope of protection of the present invention. In addition, the elements therein can be updated as technology develops.

Claims

1. UAV automatic intelligent inspection system, characterized by: It includes a UAV terminal, a ground management platform, a data transmission terminal, and an inspection and control terminal; the ground management platform is used to plan inspection tasks and flight trajectories for the UAV terminal; the UAV terminal is used to conduct inspections and collect sensor data required for the inspection tasks based on the inspection tasks and flight trajectories; The inspection and control terminal is used to monitor the status and environmental changes of the UAV terminal in real time, and to control the flight position of the UAV terminal during the process of collecting sensor data in combination with the sensor data; the data transmission terminal is used to realize sensor data transmission between the UAV terminal and the ground management platform based on 5G wireless communication technology; The data transmission terminal includes a raw data preprocessing module and a data transmission module; the raw data preprocessing module is used to preprocess the sensor data to generate target data; The data transmission module is used to transmit target data to the ground management platform.

2. The UAV automatic intelligent inspection system according to claim 1, characterized in that: The ground management platform includes an inspection task planning module, a flight trajectory planning module and a task monitoring module; the inspection task planning module is used to plan inspection tasks for the UAV terminal and determine the preset areas and target equipment that need to be inspected; the preset areas are flight areas pre-set by the administrator in the places specified by the inspection tasks; the flight trajectory planning module is used to plan the flight path and flight altitude for the UAV terminal; the task monitoring module is used to monitor the flight status of the UAV terminal and the execution of the inspection tasks in real time.

3. The automatic intelligent inspection system of a drone according to claim 2, characterized in that: The UAV terminal includes a UAV body, a flight control module, a sensor data acquisition module, and a task execution module; the flight control module, the sensor data acquisition module, and the task execution module are installed on the UAV body; the flight control module is used to adjust the flight position and flight status of the UAV body according to the inspection task, flight path, and flight altitude of the ground management platform; the task execution module is used to receive the inspection task issued by the ground management platform and determine the type of data required to be collected by the inspection task; The sensor data acquisition module is used to collect required sensor data during the inspection process.

4. The automatic intelligent inspection system of a drone according to claim 3, characterized in that: The inspection and control terminal includes a real-time status monitoring module, an environmental change perception module and a flight position control module; the real-time status monitoring module is used to monitor the flight status of the drone body in real time; the environmental change perception module is used to obtain changes in environmental data of the drone body during flight; the flight position control module is used to adjust the position of the drone body in a preset area according to changes in sensor data and environmental data.

5. The automatic intelligent inspection system of a drone as claimed in claim 4, characterized in that: The flight position control module includes a feedback control index calculation submodule and a flight position control submodule within a preset area; the feedback control index calculation submodule is used to calculate the feedback control index of the drone body currently in the preset area based on changes in sensor data and environmental data; the flight position control submodule within the preset area is used to control the flight position of the drone body within the preset area based on the situation of the preset area, the flight state of the drone body and the feedback control index; When the feedback control index calculation submodule calculates, the following formula is satisfied: Among them, G represents the feedback control index of the drone body currently in the preset area; K represents the shrinkage factor; X i represents the raw data value of the i-th sensor in the sensor data acquisition module; N represents the total number of sensor types in the sensor data acquisition module; α i represents the amplification factor of the i-th sensor; Y i represents the noise influence coefficient of the i-th sensor in the sensor data acquisition module. The higher the noise level in the raw data of the i-th sensor, the greater the corresponding noise influence coefficient; Z i Represents the reliability score of the i-th sensor in the sensor data acquisition module; Max i Represents the preset maximum value of the sum of the i-th sensor, which is used to standardize the value of the sensor sum; β i represents the nonlinear control coefficient of the i-th sensor; γ i Indicates the adjustment coefficient of the i-th sensor; Q i represents the sensitivity coefficient of the i-th sensor; w represents the weight coefficient of the product term; When G<g ref When the flight position control submodule in the preset area controls the flight position of the drone body in the preset area according to the situation of the preset area and the flight status of the drone body, the G value is optimized until G≥g ref ;g ref Indicates the evaluation threshold of the drone's flight position when the drone terminal collects sensor data.

6. The UAV automatic intelligent inspection system according to claim 5, characterized in that: The raw data preprocessing module includes a multi-type sensor data preprocessing submodule and a target data generation submodule; the multi-type sensor data preprocessing submodule is used to preprocess the raw data of various sensors; the target data generation submodule is used to package the preprocessed data into target data.

7. The automatic intelligent inspection system of a drone according to claim 6, characterized in that: The multi-type sensor data preprocessing submodule includes an image data preprocessing unit, a temperature data preprocessing unit, a vibration data preprocessing unit and a humidity data preprocessing unit; the image data preprocessing unit is used to perform gradient enhancement processing and noise reduction processing on the original data of the image sensor, and generate an image enhancement index; the temperature data preprocessing unit is used to perform deviation degree analysis on the original data of the temperature sensor, and generate a temperature global index; the vibration data preprocessing unit is used to amplify and suppress interference on the original data of the vibration sensor, and generate a vibration global index; the humidity data preprocessing unit is used to perform data smoothing and screening on the original data of the humidity sensor, and generate a humidity global index; the target data generation submodule is used to package each type of preprocessed sensor data, image enhancement index, temperature global index, vibration global index and humidity global index into target data.

8. The automatic intelligent inspection method of a drone is applied to the automatic intelligent inspection system of a drone as claimed in claim 7, characterized in that: The automatic intelligent inspection method of the drone includes: S1, planning inspection tasks and flight trajectories; S2, based on the inspection task and flight trajectory, conducts inspections and collects sensor data required by the inspection task; S3, real-time monitoring of the status of the UAV terminal and environmental changes, and combined with sensor data to control the flight position of the UAV terminal during the process of collecting sensor data; S4, based on 5G wireless communication technology, realizes sensor data transmission between drone terminals and ground management platforms.

Images