Photovoltaic field area unmanned aerial vehicle inspection positioning system

Abstract

The invention, which relates to the technical field of unmanned aerial vehicle inspection, discloses a photovoltaic field area unmanned aerial vehicle inspection positioning system comprising an unmanned aerial vehicle body, an environment sensing module, a self-adaptive flight control module, a positioning module and a data acquisition and transmission module. Through the advanced environment perception sensor and the self-adaptive flight control module, the environment condition can be accurately detected, the self-adaptive flight control module adjusts flight parameters in real time according to environment changes, damage to components of the unmanned aerial vehicle due to overheating is avoided at the extremely high temperature, normal operation of key equipment is guaranteed at the low temperature, and the safety of the unmanned aerial vehicle is guaranteed. Flight attitude and parameters can be adjusted in strong wind, stable flight is ensured not to deviate from a route, damage of sand and dust to equipment is prevented in a sand and dust environment, and a flight path is optimized, so that the unmanned aerial vehicle keeps flight stability in severe environments such as high temperature, low temperature, strong wind, sand and dust and the like; flight accidents and inspection interruption caused by severe environments are greatly reduced, and the completion rate of inspection tasks is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle inspection, and in particular to a photovoltaic field unmanned aerial vehicle inspection and positioning system. Background Art

[0002] As the global demand for clean energy grows, photovoltaic energy, as a sustainable and pollution-free energy solution, is becoming increasingly important in the energy field. The scale of photovoltaic fields is constantly expanding, from early small-scale distributed photovoltaic power stations to large-scale centralized photovoltaic power stations, covering an area of ​​several square kilometers or even larger. This poses unprecedented challenges to the equipment management and maintenance of photovoltaic fields.

[0003] The traditional inspection method of photovoltaic fields mainly relies on manual inspection. However, manual inspection has many disadvantages. On the one hand, manual inspection is inefficient. For large-area photovoltaic fields, inspectors need to spend a lot of time to complete a comprehensive inspection. Moreover, in complex terrain or severe weather conditions, the inspection work may be seriously hindered or even impossible to carry out. On the other hand, the accuracy of manual inspection is affected by the experience and working status of the personnel. Some minor faults or potential problems may be missed by the inspectors. In order to improve the efficiency and accuracy of inspections, drone inspection technology came into being. Drones can quickly cover large areas of photovoltaic fields and can collect more comprehensive and accurate photovoltaic component information by carrying various sensors.

[0004] However, the environment of photovoltaic fields is complex and changeable, and often encounters severe weather conditions such as high temperature, low temperature, strong wind, and dust. These harsh environments have a great negative impact on the flight performance and inspection functions of drones. High temperature may cause the electronic equipment of drones to overheat and be damaged, low temperature will affect battery performance and flight stability, strong wind may cause drones to deviate from the inspection route or even crash, and dust will interfere with the normal operation of sensors and data transmission. These problems seriously restrict the effective application of drone inspection technology in photovoltaic fields. Therefore, it is necessary to propose a photovoltaic field drone inspection positioning system to solve the problems in the existing technology. Summary of the invention

[0005] The purpose of the present invention is to make up for the shortcomings of the existing technology and provide a photovoltaic field drone inspection and positioning system. Through advanced environmental perception sensors and adaptive flight control modules, the drone can maintain a high degree of flight stability in harsh environments such as high temperature, low temperature, strong wind, and dust, greatly reducing flight accidents and inspection interruptions caused by harsh environments and improving the completion rate of inspection tasks.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a photovoltaic field drone inspection and positioning system, the system includes a drone body, an environment perception module, an adaptive flight control module, a positioning module and a data acquisition and transmission module, wherein:

[0007] The environmental sensing module includes a temperature sensor, a wind speed and direction sensor, and a dust sensor. The temperature sensor is used to measure the ambient temperature in the range of -50°C to 100°C with a resolution of 0.01°C. The wind speed and direction sensor is used to measure the wind speed of 0-100m / s with a resolution of 0.05m / s. The dust sensor is used to measure the air pressure of 0-1000mg / m 3 Dust concentration, resolution 0.1mg / m 3 , each sensor transmits the collected data to the adaptive flight control module;

[0008] The adaptive flight control module has a built-in real-time parameter adjustment algorithm based on fuzzy logic and neural network. The temperature value is defined as T. When T ≥ 60 ° C, the flight speed adjustment coefficient K v The relationship with temperature is K v =1-(T-60)×0.005, cooling system power adjustment coefficient K p The relationship with temperature is K p =1+(T-60)×0.008, when T<0℃, the battery heating power adjustment coefficient K h The relationship with temperature is K h =1-T×0.02, power adjustment coefficient of thermal insulation device K b The relationship with temperature is K b =1-T×0.03, for wind speed V′ w and wind direction α propeller speed adjustment coefficient K r =1+V w ×0.003, in a dusty environment, the flight altitude adjustment value H and the dust concentration C satisfy H=10+C×0.02, and the flight parameters are adjusted according to the received environmental data;

[0009] The positioning module uses the deep fusion technology of satellite positioning system and inertial navigation system, fuses data through Kalman filter algorithm, relies on inertial navigation system for short-term independent positioning in harsh environment, and uses the landmarks of photovoltaic field and characteristic information of photovoltaic modules to calibrate the positioning data based on positioning calibration algorithm;

[0010] The data acquisition and transmission module, including visible light cameras, infrared thermal imaging cameras, ultraviolet sensors, and ultrasonic sensors, is used to collect information on the appearance, temperature, material aging, spacing, and flatness of photovoltaic modules. It adopts a wireless communication module based on software-defined radio technology, combined with spread spectrum communication technology and diversity reception technology to transmit data in an anti-interference manner. The data is encrypted using the Advanced Encryption Standard.

[0011] Furthermore, in the real-time parameter adjustment algorithm in the adaptive flight control module, the specific combination of fuzzy logic and neural network is as follows:

[0012] The input layer nodes of the neural network receive the temperature, wind speed, wind direction, and dust concentration data from the environmental perception module. The middle hidden layer adopts a multi-layer perceptron structure, and the number of neurons in each layer is n1, n2, ..., n m , where m is the number of hidden layers, and the neuron activation function is

[0013] The fuzzy logic part divides temperature, wind speed, wind direction and dust concentration into different fuzzy subsets and establishes a fuzzy rule base through fuzzy reasoning rules;

[0014] The output layer of the neural network is connected to the fuzzy logic part, and the output value of the neural network is mapped to the input space of the fuzzy logic. After the fuzzy reasoning and defuzzification process, the accurate flight speed adjustment coefficient K is obtained. v , Cooling system power adjustment factor K p , Battery heating power adjustment coefficient K h , Propeller speed adjustment coefficient K r And the flight altitude adjustment value H, to achieve intelligent adjustment of the UAV flight parameters, so that the UAV can fly stably and conduct effective inspections in complex and harsh environments.

[0015] Furthermore, when the Kalman filter algorithm in the positioning module fuses the satellite positioning system and the inertial navigation system data, the specific steps are as follows:

[0016] First, define the position measured by the satellite positioning system as X s , speed is V s , the position measured by the inertial navigation system is X i , speed is V i , and set the measurement noise covariance matrix to be R s , R i , combining the position and velocity information of the satellite positioning and inertial navigation system into a system state vector X = [X s , V s , X i , V i ] T, and determine the state transfer matrix F, the control input matrix B and the control input u, and the observation matrix H;

[0017] Then, the state prediction is carried out according to the dynamic model of the system through the state prediction equation X - =FX+Bu is used to predict the system state. In this process, due to the existence of noise in the system, it is necessary to consider the system noise covariance matrix Q. The error covariance prediction equation P is estimated by prior estimation. - =FPF T +Q calculates the prior estimation error covariance P - ;

[0018] Next, calculate the Kalman gain K, which is a key parameter in the fusion process. - H T (HP - H T +R) -1 calculate;

[0019] After that, the state is updated, using the measured value Z, through the state update equation X = X - +K(Z-HX - ) Update the predicted system state to obtain a more accurate system state estimate;

[0020] Finally, update the posterior estimation error covariance by the posterior estimation error covariance update equation P = (I-KH)P - , ready for the next iteration.

[0021] Furthermore, the positioning calibration algorithm is specifically implemented based on the photovoltaic field landmarks and photovoltaic module feature information as follows:

[0022] The set of PV field landmarks and PV module feature points is defined as {P1, P2, ..., P n}, each feature point P i The coordinates in three-dimensional space are (x i ,y i , z i ), whose corresponding weight is w i , the current measurement position of the drone is (X m , Y m , Z m ), for each feature point P i , calculate its distance to the drone Constructing the objective function The drone position is calibrated by minimizing the objective function J, and the gradient descent method is used to solve the minimum value of the objective function.

[0023] Furthermore, the specific modulation and demodulation methods of the wireless communication module based on software defined radio technology in the data acquisition and transmission module during anti-interference data transmission are as follows:

[0024] In the modulation process, a hybrid modulation method is used to divide the data into two parts: high frequency band and low frequency band. For the high frequency band data D h , using multi-carrier modulation, defining the carrier frequency as f c , the number of subcarriers is N, the subcarrier spacing is Δf, then the frequency f of the kth subcarrier k =f c +kΔf (k=0,,1,…,N-1), high frequency band data D h The modulation signal Among them, A k ,φ k are the amplitude and phase of the kth subcarrier respectively;

[0025] For low frequency data D l , using pulse width modulation, low frequency data D l The modulation signal S l (t) is related to the data bit and restores the low-frequency data D by detecting the pulse width l In the demodulation process, the high-frequency band received signal is converted to the frequency domain to restore the high-frequency band data D h .

[0026] Furthermore, the specific implementation mechanism of the temperature sensor in the environment sensing module to ensure accuracy within the range of -50°C to 100°C is as follows:

[0027] The temperature sensor uses a sensing element that combines a bimetallic strip and a thermistor to measure temperature. The sensor shell adopts a double-layer insulation structure. The outer layer is a high- and low-temperature resistant ceramic material, and the inner layer is a metal material with good thermal conductivity. The middle is filled with a vacuum insulation layer to reduce the impact of the external environment on the heat conduction inside the sensor, thereby ensuring high-precision temperature measurement in a wide temperature range of -50°C to 100°C.

[0028] Furthermore, the specific measurement principle and algorithm of the wind speed and direction sensor in the environmental perception module based on the ultrasonic time difference method are as follows:

[0029] The propagation speed of ultrasound in still air is defined as c. The sensor arranges three ultrasonic transceivers A, B, and C along the wind direction. The distance between AB and AC is d. When wind blows, the wind speed and direction are calculated by measuring the ultrasonic propagation time based on the relationship between the propagation speed of ultrasound from A to B and C and the wind speed and direction. The sensor accurately measures the time through a high-precision time measurement circuit. The special wind and sand protection design and optimized installation position of the probe ensure the accuracy and reliability of the measurement, providing key environmental information for the adaptive flight control module.

[0030] Furthermore, the specific measurement algorithm of the dust sensor in the environmental perception module based on the light scattering principle and the working mechanism of the internal automatic cleaning device are as follows:

[0031] Define the particle size of dust particles as r, the scattered light intensity I and the incident light intensity I0, dust particle concentration C, particle size r, scattering angle θ and wavelength λ of light satisfy the Mie scattering theory formula Among them, a n , b n The Mie scattering coefficient is related to the refractive index, particle size and scattering angle of the dust particles. By measuring the scattered light intensity at different angles and combining the known incident light intensity and wavelength parameters with an iterative algorithm, the Mie scattering theory formula is solved to obtain the particle size distribution and concentration information of the dust particles.

[0032] The working mechanism of its internal automatic cleaning device includes: a micro fan and a cleaning brush are arranged inside the sensor. When it is detected that the dust adhesion reaches a certain level, the micro fan starts, and then the cleaning brush removes the residual dust under the drive of the motor. The bristle material of the cleaning brush can effectively absorb the dust particles. The cleaning process is carried out regularly to ensure that the dust sensor can work stably in a high dust environment.

[0033] Furthermore, the specific technical measures for ensuring the image quality of the visible light camera in the data acquisition and transmission module in harsh environments are as follows:

[0034] The visible light camera lens uses multi-layer composite optical materials, and the refractive index of each layer of material is n i and thickness d i According to the optimization of optical design theory, the incident angle of light is defined as θ i , the refraction angle is θ t , satisfying the law of refraction n i sinθ i =n t sinθ t By rationally designing the refractive index and thickness of each layer of material, the lens has high transmittance and low chromatic aberration in a wide spectral range. At the same time, the camera is equipped with a compensation system, autofocus, and anti-shake functions to ensure image quality.

[0035] Compared with the existing technology, the photovoltaic field drone inspection and positioning system has the following beneficial effects:

[0036] The present invention uses advanced environmental perception sensors and adaptive flight control modules to accurately detect environmental conditions. The adaptive flight control module adjusts flight parameters in real time according to environmental changes. Under extremely high temperatures, it prevents UAV components from being damaged due to overheating. Under low temperatures, it ensures the normal operation of key equipment. In strong winds, the flight attitude and parameters can be adjusted to ensure stable flight without deviating from the route. In dusty environments, it prevents dust from damaging equipment and optimizes the flight path. The UAV of the present invention maintains flight stability in harsh environments such as high temperature, low temperature, strong wind, and dust, greatly reducing flight accidents and inspection interruptions caused by harsh environments and improving the completion rate of inspection tasks.

[0037] Other advantages, objectives and features of the present invention will be set forth in part in the following description and, in part, will be apparent to those skilled in the art based on an examination of the following or may be taught from the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0039] Figure 1 The figure is a schematic diagram of the structure of a photovoltaic field drone inspection and positioning system. DETAILED DESCRIPTION

[0040] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] Embodiment 1

[0042] In a large photovoltaic field, the terrain is complex and it is often faced with various severe weather conditions, so the photovoltaic field drone inspection and positioning system of the present invention is deployed to ensure the normal operation of photovoltaic equipment.

[0043] First, start the drone inspection and positioning system. After the drone body is ready, the environmental perception module starts working. The temperature sensor, wind speed and direction sensor, and dust sensor are started at the same time to monitor the surrounding environment and transmit the acquired environmental data to the adaptive flight control module in real time.

[0044] The adaptive flight control module receives data from the environmental perception module and sets the temperature value to T. When the system receives the temperature information, it makes a judgment based on the real-time parameter adjustment algorithm combined with its built-in fuzzy logic and neural network. When T is in the high temperature range (assuming the high temperature threshold is T high ), the flight speed adjustment factor K v The relationship with temperature is K v =1-(TT high )×k1 (k1 is a coefficient determined according to system characteristics), cooling system power adjustment coefficient K p The relationship with temperature is K p =1+(TT high )×k2 (k2 is a coefficient determined according to system characteristics), the system reduces the flight speed and increases the power of the cooling system. When T is in the low temperature range (assuming the low temperature threshold is T low ), the battery heating power adjustment coefficient K h With Wen K b =1-T×k4 (k4 is a coefficient determined according to system characteristics), while optimizing the flight attitude.

[0045] For wind speed V w and wind direction α, propeller speed adjustment coefficient K r =1+V w ×k5 (k5 is a coefficient determined according to system characteristics), the flight attitude roll angle adjustment value β, the pitch angle adjustment value γ are calculated by specific functions based on wind speed and wind direction information. The system adjusts the propeller speed and flight attitude according to these calculation results to ensure that the drone can maintain stable flight in strong winds. When the dust data shows a dust environment, the dust concentration is set to C i The flight altitude adjustment value H and the dust concentration C satisfy H=H0+C×k6 (H0 is the basic flight altitude parameter, k6 is a coefficient determined according to the system characteristics, unit: meter). The system adjusts the flight altitude and path accordingly to avoid areas with high dust concentration.

[0046] At the same time, the positioning module is started, using the deep integration technology of satellite positioning system and inertial navigation system. Suppose the position measured by the satellite positioning system is X s , speed is V s , the position measured by the inertial navigation system is X i , speed is V i , the measurement noise covariance matrix is ​​Rs , R i , the system state vector X = [X s , V s , X i , V i ] T , state transfer matrix F, control input matrix B, control input u, observation matrix H, prior estimation error covariance P - , posterior estimation error covariance P, Kalman gain K.

[0047] set up is the predicted state estimate, which is calculated as Here X is the state vector of the previous moment, F describes the state transition law of the system from one moment to the next, B represents the influence matrix of the control input on the state, and u is the control input vector. In this process, due to the existence of noise in the system, it is necessary to consider the system noise covariance matrix Q and the prior estimation error covariance P - Calculated by: P - =FPF T +Q.

[0048] The calculation of Kalman gain K is crucial for integrating satellite positioning and inertial navigation data. The calculation formula is K = P - H T (HP - H T +R) -1 , where R is the comprehensive measurement noise covariance that comprehensively considers the measurement noise of the satellite positioning system and the inertial navigation system. After obtaining the Kalman gain K, the state is updated and the measurement value Z (combining the measurement results of the satellite positioning and inertial navigation systems) is used to update the state through the state update equation The predicted system state is updated to obtain a more accurate estimate of the system state.

[0049] Finally, update the posterior estimation error covariance by the posterior estimation error covariance update equation P = (I-KH)P - , ready for the next iteration.

[0050] Under normal circumstances, the satellite positioning system provides the main location information, and the inertial navigation system assists in calibration. When encountering severe weather such as strong sandstorms interfering with satellite signals or complex terrain such as mountainous areas causing satellite signals to be blocked, the inertial navigation system can independently provide reliable location information for the drone in a short time. In addition, by utilizing the characteristic information of landmarks and photovoltaic modules in the photovoltaic field, by constructing the objective function J (specifically constructed based on factors such as the distance between the feature points and the drone), the gradient descent method is used to solve the minimum value of the objective function to calibrate the positioning data, thereby ensuring the positioning accuracy of the drone during the entire inspection process.

[0051] The visible light camera, infrared thermal imaging camera, ultraviolet sensor, and ultrasonic sensor in the data acquisition and transmission module start working to collect information such as the appearance image, temperature distribution, material aging, component spacing and flatness of the photovoltaic components. The data collected by these sensors are transmitted through a wireless communication module based on software-defined radio technology.

[0052] In the modulation process, for the high frequency band data D h , using multi-carrier modulation, defining the carrier frequency as f c , the number of subcarriers is N, the subcarrier spacing is Δf, then the frequency f of the kth subcarrier k =f c +kΔf(k=0,1,…,N-],high frequency band data D h The modulation signal S h (t) = ∑ k =0 N-1 A k cos(2πf k t+φ k )(where A k ,φ k are the amplitude and phase of the kth subcarrier, respectively, and are determined by the high-frequency data D h Mapping obtained).

[0053] For low frequency data D l , using pulse width modulation, assuming the reference pulse width is T0, then the frequency band data D l The modulation signal (Where M is the low frequency band data length, T n Based on the low frequency data D l The pulse width determined by the nth data bit, p(t) is the unit pulse function). In the demodulation process, for the high-frequency received signal R h (t), convert it to the frequency domain through fast Fourier transform to obtain the amplitude and phase information of each subcarrier, and then restore the high frequency band data D through inverse mapping h .

[0054] For the low frequency band receiving signal R l (t), restore the low-frequency data D by detecting the pulse width l This hybrid modulation and demodulation method combines spread spectrum communication technology and diversity reception technology to resist interference in harsh environments. At the same time, the transmitted data is encrypted using advanced encryption standards to ensure the security of the data during transmission.

[0055] During the entire inspection process, the drone flies according to the planned route, continuously collects data and transmits it to the ground control center. After receiving the data, the ground control center decrypts and analyzes it, and uses the data to determine the status of the photovoltaic modules and detect potential faults in a timely manner.

[0056] Effects brought about by this embodiment: Through the application of this embodiment, in harsh environments such as high temperature, low temperature, strong wind, and dust, the UAV adjusts flight parameters based on precise calculation formulas to maintain stable flight and accurate inspection. The positioning module ensures high-precision positioning in complex environments through complex algorithms. The advanced modulation and demodulation technology of the data acquisition and transmission module ensures stable transmission of data in harsh environments. This reliable inspection capability ensures efficient maintenance of photovoltaic fields, can promptly discover problems with photovoltaic components, and reduce power generation losses caused by equipment failures. At the same time, precise positioning and secure data transmission ensure the accuracy and stability of the entire inspection system, greatly improving the quality and efficiency of photovoltaic field inspections and reducing maintenance costs and manpower investment.

[0057] Embodiment 2

[0058] In a large photovoltaic field located in a coastal area, which is often affected by the marine climate, high temperature and humidity, strong winds and rainstorms, and occasional sandstorms, the photovoltaic field drone inspection and positioning system of the present invention is deployed here to inspect photovoltaic equipment.

[0059] After starting the system, the drone is ready to take off, the environmental perception module starts to operate, the temperature sensor senses the ambient temperature T in real time, and the wind speed and wind direction sensor detects the wind speed V w and wind direction α, the dust sensor monitors the dust conditions (dust concentration is represented by C) and transmits these data to the adaptive flight control module.

[0060] After receiving the temperature data T, the adaptive flight control module analyzes it according to its internal real-time parameter adjustment algorithm based on fuzzy logic and neural network. When the temperature is in the high temperature range that will affect the normal flight of the drone, the flight speed adjustment coefficient K v The relationship with temperature is expressed as K v =1-(TT h2 )×k7(T h2 is the high temperature threshold in this scenario, k7 is a coefficient determined according to system characteristics, and the cooling system power adjustment coefficient K p The relationship with temperature is K p =1+(TT h2 )×k8 (k8 is the correlation coefficient), the system adjusts the flight speed and heat dissipation power accordingly. When in a low temperature state (T <T l2 ,T l2is the low temperature threshold in this scenario), the battery heating power adjustment coefficient K h The relationship with temperature is K h =1-T×k9 (k9 is a specific coefficient), power adjustment coefficient of the insulation device K b The relationship with temperature is K b =1-T×k 10 (k 10 is a specific coefficient) and optimizes the flight attitude at the same time.

[0061] For wind speed V w . and wind direction α, propeller speed adjustment coefficient K r =1+V w ×k 11 (k 11 is the coefficient designed according to the system), the flight attitude adjustment parameters (roll angle β, pitch angle γ) are calculated through a specific function F3 (V w ,α) and F4(V w , α) is calculated to adjust the UAV flight attitude and propeller speed to ensure stable flight in strong winds. When the dust sensor detects a dusty environment, the flight altitude adjustment value H and the dust concentration C satisfy H = H1 + C × k 12 (H1 is the initial flight altitude related parameter in this scenario, k 12 The path adjustment is optimized according to the gradient vector of the dust concentration distribution. The direction and magnitude of the path adjustment are determined by calculating the direction and magnitude of the gradient vector. The gradient vector calculation is based on the rate of change of the dust concentration data collected in real time in three-dimensional space, so as to avoid high dust areas.

[0062] The positioning module is started at the same time, which deeply integrates the satellite positioning system with the inertial navigation system. Suppose the position measured by the satellite positioning system is X s2 , speed is V s2 , the position measured by the inertial navigation system is X i2 , speed is V i2 , and the respective measurement noise covariance matrices are R s2 , R i2 , the system state vector X2 = [X s2 , V s2 , X i2 , V i2 ] T , state transfer matrix F2, control input matrix B2, control input u2, observation matrix H2, prior estimation error covariance Posterior estimation error covariance P2, Kalman gain K2.

[0063] set up is the predicted state estimate, which is calculated as Here X2 is the state vector of the previous moment, F2 describes the state transition law of the system from one moment to the next, B2 represents the influence matrix of the control input on the state, u2 is the control input vector, considering the system noise covariance matrix Q2, the prior estimation error covariance pass calculate.

[0064] The calculation formula of Kalman gain K2 is: Where R2 is the comprehensive measurement noise covariance, and the measurement value Z2 (combining the measurement results of satellite positioning and inertial navigation system) is used to update the state equation Updates are made to the predicted system status.

[0065] Finally, the error covariance is updated by the posterior estimate Update the posterior estimation error covariance to prepare for the next iteration.

[0066] Under normal circumstances, the satellite positioning system is dominant in positioning, with inertial navigation as an auxiliary. When severe weather interferes with satellite signals or complex terrain blocks signals, the inertial navigation system independently provides reliable position information for the UAV. At the same time, the landmarks and characteristic information of photovoltaic modules in the photovoltaic field are used for positioning calibration. By constructing the objective function J2 (based on factors such as the spatial relationship between the feature points and the UAV), the gradient descent method is used to solve the minimum value of the objective function to calibrate the positioning data, thereby ensuring the positioning accuracy of the UAV throughout the inspection process.

[0067] The visible light camera, infrared thermal imaging camera, ultraviolet sensor, and ultrasonic sensor of the data collection and transmission module are put into operation to collect information such as the appearance, temperature, material aging, spacing and flatness of photovoltaic modules. The collected data is transmitted through a wireless communication module based on software-defined radio technology.

[0068] In the modulation process, for the high frequency band data D h2 , using multi-carrier modulation, assuming the carrier frequency is f c2 , the number of subcarriers is N2, and the subcarrier spacing is Δf2, then the frequency f of the k2th subcarrier is k2 =f c2 +k2Δf2(k2=0,1,…,N2-1), high frequency band data D h2 The modulation signal (A k 2. φ k2 are the amplitude and phase of the k2th subcarrier respectively, which are determined by the high-frequency band data D h2 Mapping obtained).

[0069] For low frequency data D l2 , using pulse width modulation, assuming the reference pulse width is T02 , then the low-frequency data (M2 is the low frequency band data length, T n2 Based on the low frequency data D l 2, where the pulse width is determined by the n2th data bit, and p(t) is the unit pulse function). In the demodulation process, for the high-frequency received signal R h2 (t), convert it to the frequency domain through fast Fourier transform to obtain the amplitude and phase information of each subcarrier, and then restore the high frequency band data D through inverse mapping h2 .

[0070] For the low frequency band receiving signal R l2 (t), restore the low-frequency data D by detecting the pulse width l2 This wireless communication module combines spread spectrum communication technology and diversity reception technology to resist interference from harsh environments and encrypts transmitted data.

[0071] During the entire inspection process, the drone flies along a preset route, continuously collects and transmits data to the ground control center. The ground control center decrypts and analyzes the data after receiving it, and judges the status of the photovoltaic modules based on it, and detects potential faults in a timely manner.

[0072] Effects brought about by this embodiment: In such a complex and harsh climate environment in coastal areas, the UAV inspection and positioning system in this embodiment can effectively cope with it. The adaptive flight control module adjusts the flight parameters according to the environmental parameter formula, so that the UAV can fly stably and inspect accurately under harsh conditions such as high temperature, strong wind, and dust. The precise positioning module uses a fusion positioning and calibration algorithm to ensure positioning accuracy. The reliable data acquisition and transmission module ensures stable transmission and security of data in harsh environments. This series of functions ensures efficient inspection of photovoltaic fields, can promptly discover problems with photovoltaic components, reduce power generation losses caused by faults, improve inspection quality and efficiency, and reduce maintenance costs and manpower investment.

[0073] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the same elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A photovoltaic field drone inspection and positioning system, characterized in that: The system includes a drone body, an environmental perception module, an adaptive flight control module, a positioning module, and a data acquisition and transmission module, among which: The environmental sensing module includes a temperature sensor, a wind speed and direction sensor, and a dust sensor. The temperature sensor is used to measure the ambient temperature in the range of -50°C to 100°C with a resolution of 0.01°C. The wind speed and direction sensor is used to measure the wind speed of 0-100m / s with a resolution of 0.05m / s. The dust sensor is used to measure the air pressure of 0-1000mg / m 3 Dust concentration, resolution 0.1mg / m 3 , each sensor transmits the collected data to the adaptive flight control module; The adaptive flight control module has a built-in real-time parameter adjustment algorithm based on fuzzy logic and neural network. The temperature value is defined as T. When T ≥ 60 ° C, the flight speed adjustment coefficient K v The relationship with temperature is K v =1-(T-60)×0.005, cooling system power adjustment coefficient K p The relationship with temperature is K p =1+(T-60)×0.008, when T<0℃, the battery heating power adjustment coefficient K h The relationship with temperature is K h =1-T×0.02, power adjustment coefficient of thermal insulation device K b The relationship with temperature is K b =1-T×0.03, for wind speed V′ w and wind direction α propeller speed adjustment coefficient K r =1+V w ×0.003, in a dusty environment, the flight altitude adjustment value H and the dust concentration C satisfy H=10+C×0.02, and the flight parameters are adjusted according to the received environmental data; The positioning module uses the deep fusion technology of satellite positioning system and inertial navigation system, fuses data through Kalman filter algorithm, relies on inertial navigation system for short-term independent positioning in harsh environment, and uses the landmarks of photovoltaic field and characteristic information of photovoltaic modules to calibrate the positioning data based on positioning calibration algorithm; The data acquisition and transmission module, including visible light cameras, infrared thermal imaging cameras, ultraviolet sensors, and ultrasonic sensors, is used to collect information on the appearance, temperature, material aging, spacing, and flatness of photovoltaic modules. It adopts a wireless communication module based on software-defined radio technology, combined with spread spectrum communication technology and diversity reception technology to transmit data in an anti-interference manner. The data is encrypted using the Advanced Encryption Standard.

2. A photovoltaic field drone inspection and positioning system according to claim 1, characterized in that: In the real-time parameter adjustment algorithm in the adaptive flight control module, the specific combination of fuzzy logic and neural network is as follows: The input layer nodes of the neural network receive the temperature, wind speed, wind direction, and dust concentration data from the environmental perception module. The middle hidden layer adopts a multi-layer perceptron structure, and the number of neurons in each layer is n1, n2, ..., n m , where m is the number of hidden layers, and the neuron activation function is The fuzzy logic part divides temperature, wind speed, wind direction and dust concentration into different fuzzy subsets and establishes a fuzzy rule base through fuzzy reasoning rules; The output layer of the neural network is connected to the fuzzy logic part, and the output value of the neural network is mapped to the input space of the fuzzy logic. After the fuzzy reasoning and defuzzification process, the accurate flight speed adjustment coefficient K is obtained. v , Cooling system power adjustment factor K p , Battery heating power adjustment coefficient K h , propeller speed adjustment coefficient K r And the flight altitude adjustment value H, to achieve intelligent adjustment of the UAV flight parameters, so that the UAV can fly stably and conduct effective inspections in complex and harsh environments.

3. A photovoltaic field drone inspection and positioning system according to claim 1, characterized in that: The Kalman filter algorithm in the positioning module integrates the satellite positioning system and the inertial navigation system data, and the specific steps are as follows: First, define the position measured by the satellite positioning system as X s , speed is V s , the position measured by the inertial navigation system is X i , speed is V i , and set the measurement noise covariance matrix to be R s , R i , combining the position and velocity information of the satellite positioning and inertial navigation system into a system state vector X = [X s , V s , X i , V i ] T , and determine the state transfer matrix F, the control input matrix B and the control input u, and the observation matrix H; Then, the state prediction is carried out according to the dynamic model of the system through the state prediction equation X - =FX+Bu is used to predict the system state. In this process, due to the existence of noise in the system, it is necessary to consider the system noise covariance matrix Q. The error covariance prediction equation P is estimated by prior estimation. - =FPF T +Q calculates the prior estimation error covariance P - ; Next, calculate the Kalman gain K, which is a key parameter in the fusion process. - H T (HP - H T +R) -1 calculate; After that, the state is updated, using the measured value Z, through the state update equation X = X - +K(Z-HX - ) Update the predicted system state to obtain a more accurate system state estimate; Finally, update the posterior estimation error covariance by the posterior estimation error covariance update equation P = (I-KH)P - , ready for the next iteration.

4. A photovoltaic field drone inspection and positioning system according to claim 1, characterized in that: The specific implementation of the positioning calibration algorithm based on the photovoltaic field landmarks and photovoltaic module feature information is as follows: The set of PV field landmarks and PV module feature points is defined as {P1, P2, ..., P n }, the coordinates of each feature point Pi in three-dimensional space are (x i ,y i , z i ), whose corresponding weight is w i , the current measurement position of the drone is (X m , Y m , Z m ), for each feature point P i , calculate its distance to the drone Constructing the objective function The drone position is calibrated by minimizing the objective function, and the gradient descent method is used to solve the minimum value of the objective function.

5. The photovoltaic field drone inspection and positioning system according to claim 1 is characterized in that: The specific modulation and demodulation methods of the wireless communication module based on software defined radio technology in the data acquisition and transmission module during anti-interference data transmission are as follows: In the modulation process, a hybrid modulation method is used to divide the data into two parts: high frequency band and low frequency band. For the high frequency band data D h , using multi-carrier modulation, defining the carrier frequency as f c , the number of subcarriers is N, the subcarrier spacing is Δf, then the frequency f of the kth subcarrier k =f c +kΔf (k=0,1,…,N-1), high frequency band data D h The modulation signal Among them, A k , are the amplitude and phase of the kth subcarrier respectively; For low frequency data D l , using pulse width modulation, low frequency data D l The modulation signal S l (t) is related to the data bit and restores the low-frequency data D by detecting the pulse width l In the demodulation process, the high-frequency band received signal is converted to the frequency domain to restore the high-frequency band data D h .

6. The photovoltaic field drone inspection and positioning system according to claim 1 is characterized in that: The specific implementation mechanism of the temperature sensor in the environmental sensing module to ensure accuracy within the range of -50°C to 100°C is as follows: The temperature sensor uses a sensing element that combines a bimetallic strip and a thermistor to measure temperature. The sensor shell adopts a double-layer insulation structure. The outer layer is a high- and low-temperature resistant ceramic material, and the inner layer is a metal material with good thermal conductivity. The middle is filled with a vacuum insulation layer to reduce the impact of the external environment on the heat conduction inside the sensor, thereby ensuring high-precision temperature measurement in a wide temperature range of -50°C to 100°C.

7. The photovoltaic field drone inspection and positioning system according to claim 1 is characterized in that: The specific measurement principle and algorithm of the wind speed and direction sensor in the environmental perception module based on the ultrasonic time difference method are as follows: The propagation speed of ultrasound in still air is defined as c. The sensor arranges three ultrasonic transceivers A, B, and C along the wind direction. The distance between AB and AC is d. When wind blows, the wind speed and direction are calculated by measuring the ultrasonic propagation time based on the relationship between the propagation speed of ultrasound from A to B and C and the wind speed and direction. The sensor accurately measures the time through a high-precision time measurement circuit. The special wind and sand protection design and optimized installation position of the probe ensure the accuracy and reliability of the measurement, providing key environmental information for the adaptive flight control module.

8. The photovoltaic field drone inspection and positioning system according to claim 1 is characterized in that: The specific measurement algorithm of the dust sensor in the environmental perception module based on the light scattering principle and the working mechanism of the internal automatic cleaning device are as follows: Define the particle size of dust particles as r, the scattered light intensity I and the incident light intensity I0, dust particle concentration C, particle size r, scattering angle θ and wavelength λ of light satisfy the Mie scattering theory formula Among them, a n 、b n The Mie scattering coefficient is related to the refractive index, particle size and scattering angle of the dust particles. By measuring the scattered light intensity at different angles and combining the known incident light intensity and wavelength parameters with an iterative algorithm, the Mie scattering theory formula is solved to obtain the particle size distribution and concentration information of the dust particles. The working mechanism of its internal automatic cleaning device includes: a micro fan and a cleaning brush are arranged inside the sensor. When it is detected that the dust adhesion reaches a certain level, the micro fan starts, and then the cleaning brush removes the residual dust under the drive of the motor. The bristle material of the cleaning brush can effectively absorb the dust particles. The cleaning process is carried out regularly to ensure that the dust sensor can work stably in a high dust environment.

9. The photovoltaic field drone inspection and positioning system according to claim 1, characterized in that: The specific technical measures to ensure image quality of the visible light camera in the data acquisition and transmission module under harsh environments are as follows: The visible light camera lens uses multi-layer composite optical materials, and the refractive index of each layer of material is n i and thickness d i According to the optimization of optical design theory, the incident angle of light is defined as θ i , the refraction angle is θ t , satisfying the law of refraction n i sinθ i =n t sinθ t By rationally designing the refractive index and thickness of each layer of material, the lens has high transmittance and low chromatic aberration in a wide spectral range. At the same time, the camera is equipped with a compensation system, autofocus, and anti-shake functions to ensure image quality.

Images