Control method and device of inspection unmanned aerial vehicle

Abstract

The invention discloses a control method and device for an inspection unmanned aerial vehicle, and the method comprises the steps: obtaining inspection sensing data which is data collected by a sensor in the inspection process of the inspection unmanned aerial vehicle; analyzing the inspection sensing data, and determining a real-time positioning inspection state of the inspection unmanned aerial vehicle; based on the real-time positioning inspection state of the inspection unmanned aerial vehicle, voiceprint signals of the inspection unmanned aerial vehicle are extracted, and position attitude data of the inspection unmanned aerial vehicle are determined in combination with a state estimation equation of the inertial measurement unit; and adjusting the inspection path of the inspection unmanned aerial vehicle according to the position attitude data to complete the inspection task of the inspection unmanned aerial vehicle. By judging the real-time positioning inspection state of the inspection unmanned aerial vehicle, the affected condition of the inspection unmanned aerial vehicle is known in time, the position attitude data of the inspection unmanned aerial vehicle is determined according to the affected condition of the inspection unmanned aerial vehicle in combination with the voiceprint signal and the state estimation equation, the inspection path is adjusted in time, and it is guaranteed that the inspection task is accurately completed.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent inspection, and in particular relates to a control method and device for an inspection drone. Background Art

[0002] As a power grid hub, the operating quality of the substation is closely related to the safe and stable operation of the entire power grid. Therefore, regular inspections are required to ensure the safe and stable operation of its equipment.

[0003] With the rapid development of inspection drone technology, its application in the power industry is becoming more and more in-depth, especially in the inspection tasks of high-risk areas such as substations. Inspection drones have become an important tool to improve inspection efficiency and reduce manual risks. However, the complex environment of substations, such as strong electromagnetic interference and dense equipment layout, poses severe challenges to the flight stability, communication reliability and overall safety of inspection drones.

[0004] On the one hand, the strong electromagnetic interference generated by high-voltage electrical equipment often causes inspection drones to have positioning deviations, flight trajectory deviations, and even communication link interruptions. In severe cases, the inspection drones lose control or crash, and the inspection mission fails. The inspection drones may even collide with high-voltage equipment, causing equipment damage or safety accidents, affecting the normal operation of the substation and causing huge economic losses. On the other hand, factors such as complex meteorological conditions, active ionospheric and unfavorable satellite geometric distribution may cause the positioning system that relies on real-time kinematic (RTK) technology to decrease in accuracy or even fail, affecting the execution capability and safety of the inspection drone inspection mission.

[0005] At present, inspection drone systems are equipped with electromagnetic shielding materials or anti-interference equipment to reduce the interference of electromagnetic fields on navigation and communication systems to ensure the flight stability and safety of inspection drones in substation environments. However, these measures have limited effect in areas with strong electromagnetic interference, and may still cause inspection drones to lose control, increase the weight and power consumption of inspection drones, and reduce flight efficiency and endurance. Inspection path planning mostly relies on static algorithms and lacks adaptive adjustment to real-time electromagnetic interference and environmental changes, which can easily cause inspection drones to enter areas with strong electromagnetic interference, affecting inspection tasks. The emergency response of existing systems mainly relies on manual intervention. When encountering communication interruption or electromagnetic interference, operators manually control the inspection drones to make emergency landings or avoid obstacles, but this method is difficult to respond to emergencies quickly, increasing the risk of inspection drones losing control.

[0006] Therefore, the inspection UAV system still has technical bottlenecks in dealing with strong electromagnetic interference, obstacle avoidance and emergency handling. How to make the inspection UAV system adapt to the complex environment of the substation and ensure that the inspection tasks are completed efficiently and accurately is the problem that needs to be solved at present. Summary of the invention

[0007] In view of the defects in the above-mentioned prior art, the present invention provides a control method for an inspection drone, comprising:

[0008] Acquire patrol sensor data, where the patrol sensor data is data collected by sensors of the patrol drone during the patrol process;

[0009] Analyze inspection sensor data to determine the real-time positioning and inspection status of the inspection drone;

[0010] Based on the real-time positioning and inspection status of the inspection drone, the voiceprint signal of the inspection drone is extracted, and combined with the state estimation equation of the inertial measurement unit, the position and attitude data of the inspection drone is determined;

[0011] According to the position and posture data, the inspection path of the inspection drone is adjusted to complete the inspection mission of the inspection drone.

[0012] In a possible implementation, the inspection sensor data includes positioning signal data and electromagnetic signal data;

[0013] Analyze inspection sensor data to determine the real-time positioning and inspection status of the inspection drone, including:

[0014] Extract signal features from positioning signal data and give positioning error;

[0015] Extract electromagnetic features from electromagnetic signal data and give electromagnetic interference intensity;

[0016] The positioning error and the electromagnetic interference intensity are judged respectively to obtain the positioning state and the electromagnetic interference state;

[0017] Based on the positioning status and / or electromagnetic interference status, the real-time positioning and inspection status of the inspection drone is determined.

[0018] In a possible implementation, the positioning signal data includes satellite signal power, and the signal characteristics include signal strength characteristics, multipath characteristics, satellite spatial distribution characteristics, carrier phase error and delay characteristics;

[0019] Extract signal features from positioning signal data and give positioning error, including:

[0020] Obtain satellite spatial distribution characteristics, carrier phase error and delay characteristics;

[0021] According to the satellite signal power, the ideal attenuation value, actual attenuation value and additional attenuation value of the satellite signal are given;

[0022] Determine the signal strength characteristics of the satellite signal based on the ideal attenuation value, the actual attenuation value and the additional attenuation value of the satellite signal;

[0023] Analyze signal strength characteristics and give multipath characteristics;

[0024] Based on the weight of each signal feature, the signal strength feature, multipath feature, satellite spatial distribution feature, carrier phase error and delay feature are integrated to obtain the positioning error.

[0025] In a possible implementation, the satellite spatial distribution characteristic is indexed by the geometrical dilution of precision;

[0026] Set the judgment threshold of geometric precision dilution;

[0027] Based on the analysis of geometric precision factor and discrimination threshold, the influence of geometric precision factor on positioning error is measured;

[0028] The geometric precision factor is specifically expressed as:

[0029]

[0030] Where GDOP is the geometric dilution of precision, q 11 is the error of the receiver’s position coordinates in the x-axis direction, q 22 is the error of the receiver’s position coordinates in the y-axis direction, q 33 is the error of the receiver’s position coordinates in the z-axis direction, q 44 is the time error of the receiver.

[0031] In a possible implementation, the electromagnetic characteristics include magnetic field intensity characteristics, magnetic field spectrum characteristics, and magnetic field change rate characteristics;

[0032] Extract electromagnetic features from electromagnetic signal data and give electromagnetic interference intensity, including:

[0033] Based on the electromagnetic signal data, the magnetic field intensity characteristics, magnetic field spectrum characteristics and magnetic field change rate characteristics are given, wherein the magnetic field spectrum characteristics are obtained by performing fast Fourier transform on the electromagnetic field signal, and the magnetic field change rate characteristics are the rate of change of the electromagnetic field intensity over time;

[0034] Based on the weights corresponding to each electromagnetic feature, the magnetic field intensity feature, magnetic field spectrum feature and magnetic field change rate feature are integrated to obtain the electromagnetic interference intensity.

[0035] In a possible implementation, based on the real-time positioning inspection status of the inspection drone, the voiceprint signal of the inspection drone is extracted, and the position and posture data of the inspection drone is determined in combination with the state estimation equation of the inertial measurement unit, including:

[0036] When the real-time positioning inspection status is abnormal, the acoustic signal features in the voiceprint signal are extracted and the observation equation is established;

[0037] Combined with the state estimation equation of the inertial measurement unit, the real-time position and real-time attitude of the inspection UAV are analyzed to obtain the position and attitude data of the inspection UAV.

[0038] In a possible implementation, the state estimation equation of the inertial measurement unit is combined to analyze the real-time position and real-time attitude of the inspection drone to obtain the position and attitude data of the inspection drone, including:

[0039] Obtaining a parameter association equation, wherein the parameter association equation includes an association relationship between each parameter in the observation equation and each parameter in the state estimation equation;

[0040] Based on the parameter association equation, the observation equation and the state estimation equation are integrated to analyze the real-time position and attitude of the inspection UAV and determine the position and attitude data of the inspection UAV.

[0041] In a possible implementation, the method further includes:

[0042] According to the position and attitude data, a first distance between the position and attitude data and a no-fly zone is given;

[0043] Based on the preset safety range, the first distance is judged to obtain a first judgment result;

[0044] According to the first judgment result, an alarm instruction is output.

[0045] In a possible implementation, adjusting the inspection path of the inspection drone according to the position and posture data to complete the inspection task of the inspection drone includes:

[0046] Obtain obstacle position data corresponding to the position and posture data;

[0047] Adjust the inspection path based on obstacle location data.

[0048] The present invention also discloses a control device for an inspection drone, comprising:

[0049] A data acquisition module is used to acquire patrol sensor data, wherein the patrol sensor data is data collected by sensors of the patrol drone during the patrol process;

[0050] The status determination module is used to analyze the inspection sensor data and determine the real-time positioning inspection status of the inspection drone;

[0051] The position analysis module is used to extract the voiceprint signal of the inspection drone based on the real-time positioning inspection status of the inspection drone, and determine the position and attitude data of the inspection drone in combination with the state estimation equation of the inertial measurement unit;

[0052] The path adjustment module is used to adjust the inspection path of the inspection drone according to the position and posture data to complete the inspection task of the inspection drone.

[0053] The control method of the inspection drone provided by the present invention has at least the following beneficial effects:

[0054] By acquiring inspection sensor data in real time, the real-time positioning and inspection status of the inspection UAV can be judged, and the impact of various environmental factors on the inspection UAV during the inspection process can be timely understood. According to the impact of the inspection UAV, the position and posture data of the inspection UAV is determined in combination with the voiceprint signal and IMU, and the inspection path is adjusted in time to ensure the accurate completion of the inspection task. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A flow chart of a control method for an inspection drone provided by an embodiment of the present invention;

[0056] Figure 2 A schematic diagram of a flow chart of a control method for an inspection drone provided by an embodiment of the present invention;

[0057] Figure 3 A structural diagram of a control device for an inspection drone provided in an embodiment of the present invention.

[0058] Among them, 201 is a data acquisition module; 202 is a state determination module; 203 is a location analysis module; 204 is a path adjustment module. DETAILED DESCRIPTION

[0059] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the drawings and specific implementation methods of the specification. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0060] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The singular forms "a", "said" and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings, and "multiple" generally includes at least two.

[0061] It should also be noted that the term "includes", "comprising" or any other variation thereof is intended to cover non-exclusive inclusion, so that a commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprising a ..." do not exclude the existence of other identical elements in the commodity or device including the elements.

[0062] With the advancement of science and technology and the continuous deepening and development of power system reform, the degree of automation of the power system has been greatly improved. The application of inspection drone technology can well complete the tasks of power inspection and construction planning.

[0063] However, the layout of equipment in the substation is complex, with densely distributed high-voltage cables, transformers, switches and other equipment. The surface of some equipment is highly reflective or has no obvious characteristics of metal materials, which increases the difficulty of obstacle avoidance for inspection drones. Inspection drones are prone to collision with these high-voltage equipment during inspections, posing serious safety hazards. In addition, the existing inspection drone system lacks effective emergency handling capabilities, making it difficult to quickly generate emergency flight routes or take emergency obstacle avoidance measures when encountering electromagnetic interference, communication interruption or aircraft loss of control.

[0064] To solve the problem of equipment obstacle avoidance, existing systems use airborne visual sensors or radar systems to identify obstacles. However, in the case of strong electromagnetic interference, low illumination, strong light, backlight, weak texture or bad weather, the system's recognition ability and reaction speed are insufficient, resulting in obstacle avoidance failure and increasing the risk of collision between inspection drones and high-voltage equipment. Although existing technologies combine inertial navigation systems (INS) with global positioning systems (GPS) to improve positioning accuracy, the problem of strong electromagnetic interference affecting GPS signals and affecting positioning accuracy has not yet been solved. How to enable inspection drones to avoid electromagnetic interference and equipment obstacles in complex environments at the same time, improve system reliability, and ensure the normal operation of the power system is a problem that needs to be solved at present.

[0065] The present invention provides a control method for an inspection drone, which obtains various sensor data of the inspection drone in real time, including RTK signals, inertial measurement unit (IMU) attitude sensor data, voiceprint information received by a microphone array, and measurement data of an electromagnetic field detector. Then the collected data is analyzed, especially the failure state of the RTK signal and the electromagnetic interference intensity are analyzed, judged and monitored. When the system detects that the RTK fails or the electromagnetic interference is too high, the emergency warning mechanism is immediately triggered. Then, the voiceprint three-dimensional positioning and ranging technology is used to provide more accurate inspection drone positioning and environmental depth information. Voiceprint three-dimensional positioning and ranging technology can effectively overcome the limitations of electromagnetic interference on traditional navigation systems, ensuring that the inspection drone can still be accurately positioned in complex environments. At the same time, the system enters a monitoring and early warning mode, and judges the distance between the inspection drone and the power equipment and the no-fly zone in real time. If the distance to the equipment or the no-fly zone is too close, a warning is issued to prevent the inspection drone from colliding with high-voltage equipment or entering the no-fly zone by mistake. Finally, in emergency situations such as RTK failure or communication interruption, the system promptly adjusts the inspection route based on voiceprint positioning, ranging information and other sensor data to ensure that the inspection drone can safely avoid high-intensity electromagnetic interference areas, power equipment and no-fly zones, thereby ensuring that the inspection drone can stably and safely complete the inspection task in a complex electromagnetic environment.

[0066] like Figure 1 As shown, the present invention provides a control method for an inspection drone, and the specific steps are as follows:

[0067] S101: Acquire inspection sensor data.

[0068] Specifically, the inspection sensor data is data collected by sensors of the inspection drone during the inspection process. The inspection sensor data includes positioning signal data and electromagnetic signal data, and the positioning signal data includes satellite signal power.

[0069] It is understandable that a variety of sensors are integrated on the inspection drone, including RTK sensors, IMU sensors, voiceprint sensors including speaker arrays and microphone arrays, and electromagnetic field detectors. The sensors on the inspection drone obtain inspection sensor data in real time, providing accurate real-time information for the flight control, obstacle avoidance and emergency handling of the inspection drone. In the embodiment provided by the present invention, the positioning signal data is the RTK signal collected by the above-mentioned RTK. The RTK sensor uses a global navigation satellite system (GNSS) receiver and a differential data processor, including more than ten data observation values ​​such as differential signals, carrier phase, pseudorange, satellite ephemeris, ephemeris data, positioning, and geodetic coordinate information. The electromagnetic signal data is the electromagnetic field data around the inspection drone measured by the electromagnetic field detector. Key electromagnetic features such as electromagnetic field intensity, electromagnetic field change rate, electromagnetic field frequency, and electromagnetic field spatial distribution can be extracted from the electromagnetic signal data.

[0070] In addition, the IMU sensor in the inspection drone can monitor the three-dimensional posture of the inspection drone, such as angular velocity, acceleration, pitch angle, roll angle, and heading angle, in real time through accelerometers, gyroscopes, and magnetometers, providing real-time feedback for flight control and ensuring flight stability.

[0071] At the same time, the speaker array in the voiceprint sensor transmits voiceprint signals of specific frequency and intensity in an omnidirectional manner through multiple speakers distributed around the inspection drone. The microphone array captures the voiceprint signals in the environment where the inspection drone is located through multiple highly sensitive microphones. By measuring the transmission and reflection time of the voiceprint signal, the inspection drone's position and distance from obstacles are accurately calculated, and the depth of the environment is perceived.

[0072] S102: Analyze the inspection sensor data to determine the real-time positioning inspection status of the inspection drone.

[0073] Specifically, the signal features in the positioning signal data are extracted to give the positioning error, which is the current positioning error of the inspection drone. The electromagnetic features in the electromagnetic signal data are extracted to give the electromagnetic interference intensity, which is the electromagnetic interference intensity to which the inspection drone is subjected in the current environment. The positioning error and the electromagnetic interference intensity are judged respectively to obtain the positioning state and the electromagnetic interference state, thereby determining the real-time positioning and inspection state of the inspection drone.

[0074] Furthermore, the above signal characteristics include signal strength characteristics, multipath characteristics, satellite spatial distribution characteristics, carrier phase error and delay characteristics. According to the satellite signal power, the ideal attenuation value, actual attenuation value and additional attenuation value of the satellite signal are given, and the signal strength characteristics of the satellite signal are determined. The signal strength characteristics are analyzed to give the multipath characteristics. Based on the weights of each signal characteristic, the signal strength characteristics, multipath characteristics, satellite spatial distribution characteristics, carrier phase error and delay characteristics are integrated to obtain the positioning error.

[0075] It is understandable that since the satellite signal strength is affected by environmental factors such as satellite position, occlusion, and reflection, the mean μ and / or variance σ of signal attenuation are used to describe the signal strength characteristics under different environments. The specific calculation steps are as follows:

[0076] First, determine the satellite signal power. The receiver in the inspection drone measures the signal strength data of multiple satellites in real time, such as the signal-to-noise ratio (SNR) and received signal strength indication (RSSI), and calculates the satellite signal power of the received signal:

[0077]

[0078] Among them, P ri is the satellite signal power of the i-th satellite received by the receiver in the inspection drone, in decibels, P noise is the internal noise power of the receiver in the inspection UAV, S ri It is the signal strength value received by the receiver in the inspection drone.

[0079] Then, the ideal attenuation value, actual attenuation value and additional attenuation value of the satellite signal are determined.

[0080] Among them, based on the satellite coordinates of the satellite and the position coordinates of the receiver in the inspection drone, the geometric distance between the satellite and the receiver is given; according to the frequency of the satellite signal, the geometric distance between the satellite and the receiver and the speed of light, the ideal attenuation value of the satellite signal is determined, which is specifically expressed as:

[0081]

[0082] in, is the ideal attenuation value of the satellite signal from the i-th satellite to the receiver in the inspection drone, in decibels, f i is the frequency of the satellite signal of the ith satellite, in Hertz, and c is the speed of light, which is approximately 3×10 8 m / s, d iis the geometric distance from the i-th satellite to the receiver in the inspection drone, in meters, (x i ,y i ,z i ) is the satellite coordinates of the ith satellite, the satellite coordinates are the satellite position coordinates obtained through the satellite ephemeris, (x r ,y r ,z r ) is the position coordinate of the receiver in the inspection drone estimated by RTK differential. In the example provided by the present invention, the ideal attenuation value is the ideal attenuation value of the satellite signal to the receiver calculated by the free space path loss model, which is a theoretical value based on distance and frequency and is applicable to an obstacle-free free space environment.

[0083] The actual attenuation value of the satellite signal is the difference between the satellite transmission power and the satellite signal power, which is specifically expressed as:

[0084] L i =P ti -P ri

[0085] Among them, L i is the actual attenuation value of the satellite signal of the i-th satellite, P ti is the satellite transmission power of the i-th satellite, P ti Set by the system, usually around 20 dBW, P ri is the satellite signal power of the i-th satellite received by the receiver in the inspection drone. Since the signal attenuation in actual scenarios also needs to consider the additional signal attenuation caused by factors such as the ionosphere, troposphere, multipath effect, and obstacle shielding, the actual signal attenuation value is usually higher than the free space loss. In the example provided by the present invention, the actual attenuation value is calculated by the difference between the satellite transmit power and the satellite signal power.

[0086] The additional attenuation value of the satellite signal is the difference between the actual attenuation value and the theoretical attenuation value, which is specifically expressed as:

[0087]

[0088] Among them, L i ′ is the additional attenuation value of the satellite signal of the i-th satellite, L i is the actual attenuation value, is the theoretical attenuation value. The additional attenuation value is the signal attenuation caused by factors such as shielding and multipath effects.

[0089] Finally, the signal strength characteristics are determined.

[0090]

[0091]

[0092] Where μ is the mean signal attenuation under different environments, N is the number of satellites measured, and σ 2 is the signal attenuation variance under different environments, is the normal distribution model of satellite signal attenuation.

[0093] By measuring the additional attenuation value and mean under different environments, the signal attenuation variance is calculated to describe the volatility and stability of the signal. 2 , the normal distribution model of signal attenuation is obtained, which is used to describe the signal strength characteristics under different environments.

[0094] Differentiate signal strength characteristics under different environments based on the normal distribution model:

[0095] In a heavily obstructed environment, the additional attenuation of most satellite signals is between 18 dB and 23 dB. The additional attenuation is large, the signal attenuation mean μ is high, and the signal attenuation variance σ is 2 In a lightly blocked environment, the additional attenuation of most satellite signals is between 5 dB and 10 dB, the additional attenuation is small, the signal attenuation mean μ is low, and the signal attenuation variance σ is 2 In an open environment, the additional attenuation value of the satellite signal is close to 0, and the signal strength is high and stable. In the example provided by the present invention, the signal attenuation mean or signal attenuation variance is used as the signal strength feature to analyze the environmental shielding situation.

[0096] Furthermore, the multipath feature is extracted, wherein the multipath feature is the fluctuation of the signal strength data collected by the receiver in a unit time, which is specifically expressed as:

[0097] S=[S1,S2,S3,…,S n ]

[0098]

[0099] Among them, S is a data set composed of n signal strength data collected by the receiver in the inspection drone within one second, S i The signal strength data measured by the receiver in the inspection drone at the i-th collection within one second, μ' is the mean of the signal strength data within one second, and β is the standard deviation of the signal strength data within one second.

[0100] It is important to understand that the standard deviation of the signal strength data within one second can reflect the degree of multipath interference. The standard deviation β can be used to quantify the fluctuation of the signal strength data in a short period of time. The larger the standard deviation of the signal fluctuation, the stronger the multipath interference. As the degree of obstruction increases, the multipath effect increases, and the fluctuation coefficient β also increases. Therefore, the multipath characteristic is the fluctuation coefficient β.

[0101] Further, the satellite spatial distribution characteristics are extracted, wherein the satellite spatial distribution characteristics are based on the geometric dilution of precision (GDOP) as an indicator;

[0102] Set the judgment threshold of geometric precision dilution;

[0103] Based on the analysis of geometric precision factor and discrimination threshold, the influence of geometric precision factor on positioning error is measured;

[0104] The geometric precision factor is specifically expressed as:

[0105]

[0106] Where GDOP is the geometric dilution of precision, q 11 is the error of the receiver’s position coordinates in the x-axis direction, q 22 is the error of the receiver’s position coordinates in the y-axis direction, q 33 is the error of the receiver’s position coordinates in the z-axis direction, q 44 is the time error of the receiver.

[0107] It is important to understand that the dilution of precision (DOP) can measure the impact of satellite geometric distribution on positioning accuracy. The smaller the DOP value, the better the satellite geometric distribution and the higher the positioning accuracy. When the DOP value exceeds the set discrimination threshold, for example, the discrimination threshold is set to 6, the satellite geometric distribution is poor, the positioning accuracy is significantly reduced, and RTK may fail. The determination of DOP is based on the geometric relationship between the satellite and the receiver, and is calculated using the coordinate matrix of the two.

[0108] The geometric precision coefficient is determined by the following steps:

[0109] First, the observation matrix G is constructed using the coordinates of visible satellites and receivers.

[0110]

[0111]

[0112] Among them, G is the observation matrix of n visible satellites, g i is the unit vector of the ith satellite relative to the receiver, (x i ,y i ,z i ) is the position coordinate of the i-th satellite, (x r ,y r ,z r ) is the position coordinate of the receiver, di is the geometric distance from the i-th satellite to the receiver in the inspection UAV, and n is the maximum value of i.

[0113] Q=(G T G) -1

[0114] Among them, Q is the covariance matrix, G T is the transposed matrix of the observation matrix G, (G T G) -1 It's G T The inverse matrix of G.

[0115] Then, the DOP value is extracted from the covariance matrix. The covariance matrix contains the precision information of position and time, and different types of DOP values ​​can be extracted from it. In this example, the geometric precision factor is extracted from the diagonal elements of the covariance matrix. In other implementations, other types of DOP values ​​can be extracted according to actual needs, such as horizontal dilution of precision (HDOP), horizontal and time dilution of precision (HTDOP), position dilution of precision (PDOP), time dilution of precision (TDOP), and vertical dilution of precision (VDOP).

[0116] In other implementations, the invisible satellite ratio may be used as an indicator of satellite spatial distribution characteristics, specifically expressed as:

[0117] N'=N total -N

[0118]

[0119] Where N′ is the number of invisible satellites, N total is the total number of satellites, N is the number of satellites visible to the receiver at the current moment, and ξ is the proportion of invisible satellites. The total number of satellites is obtained through the ephemeris data of the receiver, and the number of visible satellites is directly measured by the receiver, which is the number of satellites that are actually visible to the receiver at the current moment and whose signal strength exceeds the set signal threshold.

[0120] It should be understood that the invisible satellite ratio can evaluate the impact of the environment on the positioning system. The invisible satellite ratio is the ratio of satellites whose signals cannot be received by the receiver due to environmental factors (such as obstruction, terrain or buildings, etc.) to the total number of satellites.

[0121] The invisible satellite ratio and geometric precision factor reflect the current satellite space geometric distribution characteristics. The overall distribution of the invisible satellite ratio and geometric precision factor is similar. In the heavily obscured area, the number of visible satellites decreases, the invisible satellite ratio and geometric precision factor increase significantly, and the positioning accuracy decreases.

[0122] Further, the carrier phase error is extracted, wherein the measured pseudorange is determined based on the satellite clock bias, the receiver clock bias, the satellite ionospheric delay and the satellite tropospheric delay; the pseudorange error is given according to the difference between the measured pseudorange and the geometric distance from the satellite to the receiver; the measured carrier phase and the theoretical carrier phase are determined in combination with the carrier wavelength to obtain the carrier phase error; specifically expressed as:

[0123] ΔD i =D i -d i

[0124] Where, ΔD i is the pseudorange error, D i To measure the pseudorange, the measured pseudorange is the actual measured distance between the receiver and the i-th satellite, including receiver clock error, ionospheric delay, tropospheric delay, multipath effect and other error terms, d i is the theoretical pseudorange, which is the geometric distance from the i-th satellite to the receiver in the inspection UAV.

[0125] D i =d i +c(Δt+Δt i )+I i +T i +ε i

[0126] Where Δt is the receiver clock deviation, Δt i is the clock bias of the ith satellite, I i is the ionospheric delay of the i-th satellite, T i is the tropospheric delay of the i-th satellite, ε i is other error terms, including noise, multipath effect, etc. Other error terms can be reduced by suppressing multipath effect and noise reduction. In the example provided by the present invention, the above ionospheric delay is obtained by dual-frequency observation or model estimation, and the above tropospheric delay is obtained by model or correction data estimation.

[0127]

[0128]

[0129] Where ΔΦ is the carrier phase error, ΔΦ is the measured carrier phase Φi Theoretical carrier phase The difference between λ and D is the carrier wavelength. i is the measured pseudorange, N is the phase integer ambiguity, d i is the theoretical pseudorange.

[0130] Integer ambiguity resolution is the key to RTK positioning accuracy. Improper resolution will lead to large phase errors, reduce positioning accuracy, and even cause RTK failure. At the same time, RTK positioning depends on the stability of ambiguity resolution. Unstable resolution will directly lead to carrier phase error jumps. The carrier phase error is detected by preset thresholds to see if it exceeds a reasonable range. When the error exceeds a certain threshold, it can be judged that RTK positioning has entered an unstable state, indicating that positioning errors may occur.

[0131] The carrier phase error evaluates the difference between the measured value and the theoretical value in the RTK system, helps analyze the GNSS measurement error, and determines whether the RTK positioning accuracy has decreased.

[0132] Furthermore, the delay features are extracted:

[0133] A d =t cur -t last

[0134] Among them, A d is the delay characteristic, t cur is the current time, t last The time when the last valid differential signal was received.

[0135] RTK systems rely on the timeliness of differential signals. The delay characteristic represents the delay of differential correction data, that is, the difference between the current reception time and the last valid differential signal reception time. The greater the delay, the worse the timeliness of the data, resulting in reduced positioning accuracy, especially in dynamic environments. The differential data delay is calculated by the difference between the base station's transmission time and the rover's reception time. When the timeliness exceeds the set threshold (such as 10 seconds), it means that the differential data is invalid, affecting the RTK accuracy.

[0136] Furthermore, based on the weights of various signal features, the signal strength features, multipath features, satellite spatial distribution features, carrier phase error and delay features are integrated to obtain the positioning error.

[0137] Among them, the signal strength characteristics, multipath characteristics, satellite spatial distribution characteristics, carrier phase error and time delay characteristics are normalized respectively; combined with the weights of each signal feature, the normalized signal strength characteristics, multipath characteristics, satellite spatial distribution characteristics, carrier phase error and time delay characteristics are fused to obtain the positioning error.

[0138] Taking into account the impact of each feature, the above features are fused to make a comprehensive judgment on the decline or failure of RTK positioning accuracy. Since the dimensions of each feature are different, for example, the signal strength feature is expressed in decibels, the DOP value is dimensionless, and the carrier phase error is expressed in meters, the features need to be normalized first to fall into the same numerical range (usually normalized to [0,1]), and then weighted fusion is performed according to specific weights.

[0139]

[0140] where x is the original eigenvalue, is the mean of the feature, σ is the standard deviation of the feature, and x′ is the standardized feature value.

[0141] F=w S ·S′+w M ·M′+w D ·D′+w Φ ·Φ′+w T ·T′

[0142] Among them, F is the positioning error, S′ is the normalized signal strength feature, M′ is the normalized multipath feature, D′ is the normalized satellite spatial distribution feature, Φ′ is the normalized carrier phase error, T′ is the normalized delay feature, and w S is the weight corresponding to the signal strength feature, w M is the weight corresponding to the multipath feature, w D is the weight corresponding to the satellite spatial distribution characteristics, w Φ is the weight corresponding to the carrier phase error, w T is the weight corresponding to the delay feature, w S It reflects the influence of signal strength characteristics on RTK positioning accuracy. M It reflects the influence of multipath characteristics on RTK positioning accuracy. D It reflects the influence of satellite spatial distribution characteristics on RTK positioning accuracy. Φ It reflects the influence of carrier phase error on RTK positioning accuracy. T It reflects the influence of delay characteristics on RTK positioning accuracy. The above weights can be set based on historical data or experience and are not limited to this.

[0143] Furthermore, the above electromagnetic characteristics include magnetic field intensity characteristics, magnetic field spectrum characteristics and magnetic field change rate characteristics. Based on the electromagnetic signal data, the magnetic field intensity characteristics, magnetic field spectrum characteristics and magnetic field change rate characteristics are given, wherein the magnetic field spectrum characteristics are obtained by fast Fourier transforming the electromagnetic field signal, and the magnetic field change rate characteristics are the rate of change of the electromagnetic field intensity over time.

[0144]

[0145]

[0146] Where E is the magnetic field intensity characteristic, E x ,E y ,E z are the intensity components of the electromagnetic field in the x, y, and z directions respectively. X(f) is the magnetic field spectrum characteristic, which is obtained by fast Fourier transforming the electromagnetic field signal. E(i) is the electromagnetic field intensity data of the ith sample, n is the number of sampling points, f′ is the electromagnetic frequency, j is the imaginary unit, ΔE is the magnetic field change rate characteristic, and the magnetic field change rate characteristic is the rate of change of the electromagnetic field intensity over time.

[0147] Furthermore, based on the weights corresponding to the various electromagnetic features, the magnetic field intensity features, the magnetic field spectrum features and the magnetic field change rate features are fused to obtain the electromagnetic interference intensity. The magnetic field intensity features, the magnetic field spectrum features and the magnetic field change rate features are normalized respectively; combined with the weights of the various electromagnetic features, the normalized magnetic field intensity features, the magnetic field spectrum features and the magnetic field change rate features are fused to obtain the electromagnetic interference intensity.

[0148] Since the dimensions of the magnetic field intensity feature, magnetic field spectrum feature, and magnetic field change rate feature are different, these features are normalized to be in the range [0,1]. After normalization, feature fusion is performed to obtain the electromagnetic interference intensity.

[0149]

[0150] Among them, F′ is the electromagnetic interference intensity, w1 is the weight corresponding to the magnetic field intensity feature, w2 is the weight corresponding to the magnetic field change rate feature, w3 is the weight corresponding to the magnetic field spectrum feature, E′ is the magnetic field intensity feature after normalization, ΔE' is the magnetic field change rate feature after normalization, and |X(f)-f|' is the magnetic field spectrum feature after normalization.

[0151] Further, when at least one of the positioning state and / or the electromagnetic interference state is an abnormal state, the real-time positioning inspection state is determined as an abnormal state.

[0152] The abnormal state of the positioning state is determined by:

[0153] Calculate the positioning error in real time and compare it with the set threshold to determine whether the RTK system fails, that is, to determine whether the positioning status is an abnormal state. If the positioning error continuously approaches or exceeds the threshold, the RTK positioning may be on the verge of reduced accuracy or failure. To improve the sensitivity of early warning, in this example, two positioning thresholds F1 and F2 are set, where 0 ≤ F1 < F2. Determine whether the RTK fails according to the following conditions:

[0154] When F < F1, it means that the positioning accuracy of the RTK system is within the acceptable range, that is, the positioning status is a normal state;

[0155] When F1 ≤ F < F2, it means that the positioning accuracy of the RTK system begins to decline but has not completely failed, that is, the positioning status is a warning state;

[0156] When F ≥ F2, it means that the RTK positioning accuracy has dropped significantly and may enter the failure state, that is, the positioning status is an abnormal state.

[0157] The setting of the above positioning thresholds F1 and F2 can be based on experimental data. Determine reasonable thresholds by simulating the operation of the RTK system in various environments, or other methods can also be used for setting, which is not limited here.

[0158] When the positioning error exceeds the threshold F1 but does not reach the threshold F2, the system triggers the warning state to remind the user that the RTK positioning accuracy begins to decline and make emergency preparations.

[0159] When the positioning error exceeds the threshold F2, the system triggers the abnormal state, indicating that the RTK positioning is close to failure, and immediately switches to the emergency positioning mode.

[0160] When the positioning error exceeds the threshold F1 and / or F2, send real-time warnings to the user in the following ways:

[0161] Sound alarm:发出声音提醒,提醒操作员注意RTK定位状态。通过巡检无人机扬声器或控制终端发出声音提醒,提醒操作员注意RTK定位状态。

[0162] Visual alarm:显示预警信息,通过颜色变化或弹窗提示操作员,定位状态用不同颜色表示,如预警状态为黄色,异常状态为红色。在巡检无人机控制界面显示预警信息,通过颜色变化或弹窗提示操作员,定位状态用不同颜色表示,如预警状态为黄色,异常状态为红色。

[0163] Remote notification:将预警信息发送到巡检无人机操作员的终端设备,如手机、平板或计算机,确保即使操作员不在现场也能及时响应。通过无线通信或网络将预警信息发送到巡检无人机操作员的终端设备,如手机、平板或计算机,确保即使操作员不在现场也能及时响应。

[0164] The abnormal state of the electromagnetic interference status is determined by the following methods:

[0165] It should be noted that in the translation of the content in item , there is an incorrect expression in the original Chinese. The correct translation should be "Sound alarm: Send a sound reminder through the loudspeaker of the inspection drone or the control terminal to remind the operator to pay attention to the RTK positioning status."; in the translation of the content in item , there is an incorrect expression in the original Chinese. The correct translation should be "Visual alarm: Display the warning information on the control interface of the inspection drone and prompt the operator through color changes or pop-up windows. The positioning status is represented by different colors. For example, the warning state is yellow and the abnormal state is red."; in the translation of the content in item , there is an incorrect expression in the original Chinese. The correct translation should be "Remote notification: Send the warning information to the terminal devices of the inspection drone operator, such as mobile phones, tablets or computers, to ensure that the operator can respond in time even if they are not on site."Compare the electromagnetic interference intensity with the set electromagnetic threshold to determine whether the electromagnetic environment is abnormal. At the same time, associate the electromagnetic interference intensity with the positioning status to comprehensively determine whether there is electromagnetic interference. To improve the sensitivity of early warning, in the example provided by the present invention, two electromagnetic thresholds F1' and F2' are set, where 0 ≤ F1' < F2'. The judgment conditions are as follows:

[0166] When F < F1', it indicates that the electromagnetic field is at a normal level and there is no electromagnetic field interference, that is, the electromagnetic interference state is a normal state.

[0167] When F1' ≤ F < F2', it indicates that the electromagnetic field exceeds the normal level and electromagnetic anomaly early warning is required. At this time, if the RTK accuracy decreases or fails (that is, the positioning status is in an early warning state or an abnormal state), it means that there is electromagnetic interference, and an early warning against electromagnetic interference needs to be issued and emergency preparations need to be made, that is, the electromagnetic interference state is an early warning state.

[0168] When F ≥ F2, it indicates that the electromagnetic field far exceeds the normal level and at the same time the RTK accuracy decreases or fails, which means that there is electromagnetic interference, that is, the electromagnetic interference state is an abnormal state.

[0169] The above settings of the electromagnetic thresholds F1' and F2' can be based on experimental data, and reasonable thresholds can be determined by simulating the interference of electromagnetic intensities in various environments. Other methods can also be used for setting, which is not limited herein.

[0170] When the electromagnetic interference intensity exceeds the threshold F1 and / or F2, real-time early warning can be sent to the user through the above-mentioned sound alarm, visual alarm, remote notification, etc., which is not limited herein.

[0171] S103: Based on the real-time positioning and inspection status of the inspection drone, extract the voiceprint signal of the inspection drone, and combine it with the state estimation equation of the inertial measurement unit to determine the position and attitude data of the inspection drone.

[0172] Specifically, when the real-time positioning and inspection status is in an abnormal state, that is, when the RTK fails and / or there is an electromagnetic anomaly, extract the acoustic signal characteristics in the voiceprint signal and establish an observation equation. Combine the state estimation equation of the inertial measurement unit to analyze the real-time position and real-time attitude of the inspection drone to obtain the position and attitude data of the inspection drone.

[0173] Furthermore, obtain a parameter correlation equation, where the parameter correlation equation includes the correlation relationships of the various parameters in the observation equation and the various parameters in the state estimation equation. Based on the parameter correlation equation, fuse the observation equation and the state estimation equation to analyze the real-time position and real-time attitude of the inspection drone to determine the position and attitude data of the inspection drone.

[0174] It should be understood that in the inspection drone system equipped with a microphone array and an IMU, the IMU has a high refresh rate but is prone to cumulative errors, while the microphone array has a higher accuracy but a low refresh rate. Therefore, when the real-time positioning inspection state is abnormal, the present invention adopts a patrol drone pose estimation algorithm that integrates voiceprint and IMU to obtain the position and attitude data of the patrol drone. IMU pre-integration establishes motion constraints, and the microphone array provides distance measurement to solve the problem that the positioning of the patrol drone is easily disturbed by strong electromagnetic, multipath, occlusion and other factors during flight. At the same time, the robust estimation of the six-dimensional pose of the patrol drone is realized under the Kalman filter framework, and the state space matrix formed by the double difference of space and time is further obtained to obtain the center of mass velocity and angular velocity of the patrol drone, thereby improving the accuracy of dynamic pose estimation. The specific steps are as follows:

[0175] For the fusion positioning system of microphone array and IMU, the posture state of the inspection drone, the characteristics of the acoustic signal and the IMU parameters are comprehensively considered to set the state vector of the filter:

[0176] x=[p,v,q,l i ,b a ,b w ] T

[0177] Among them, x is the state vector of the filter, p is the center of mass position of the inspection drone, v is the speed of the inspection drone, q is the posture of the inspection drone, and l i is the distance from the center of mass of the inspection drone to the i-th base station, b a is the zero bias parameter of the accelerometer in the IMU, b w is the zero bias parameter of the gyroscope in the IMU.

[0178] The fusion positioning system uses IMU to predict the center of mass motion of the inspection drone, establishes a kinematic model, and obtains the state estimation equation of the IMU:

[0179]

[0180] in, is the rate of change of the center of mass position of the inspection drone, v is the speed of the inspection drone, is the velocity change rate of the inspection UAV’s center of mass, R T is the transpose of the rotation matrix, R T It is used to convert the acceleration from the body coordinate system of the inspection drone to the IMU coordinate system where the IMU is located. a is the acceleration measured by the IMU, and n is a is the acceleration measurement noise, g is the gravity acceleration vector, which is used to compensate for the influence of gravity on the acceleration of the inspection drone. is the attitude change rate of the inspection UAV, ω is the angular velocity measured by IMU, n wis the angular velocity measurement noise, and q is the attitude of the inspection UAV.

[0181] The microphone array and IMU fusion positioning system obtains distance measurement by extracting acoustic signal features, and uses double differential technology to jointly estimate the position and attitude of the inspection drone. The differential equations of different microphones in the same base station are used for attitude estimation, and the differential equations between different base stations of the same microphone are used for position estimation, and then the observation equation is established. Among them, the acoustic signal characteristics include the speed of sound, the time when the microphone array receives the base station acoustic signal, the time when the base station transmits the acoustic signal, the time delay caused by the base station transmission and the microphone reception, and the correction factor of the acoustic signal. Based on the acoustic signal characteristics, the observation equation is established, which is specifically expressed as:

[0182] l i =v sound (t i -t0-iΔt)+ζ i

[0183]

[0184] Among them, l i is the distance from the center of mass of the inspection drone to the i-th base station, v sound is the speed of sound, v sound =343m / s,t i is the time when the microphone array receives the sound signal from the i-th base station, t0 is the time when the i-th base station transmits the sound signal, iΔt is the time delay between the transmission of the i-th base station and the reception of the microphone, ζ i is the correction factor of the ith sound signal, t last is the last measurement time, T is the time period, is the relative velocity vector between the ith base station and the inspection drone.

[0185] In addition, the parameters in the observation equation and the parameters in the state estimation equation are related, which can be specifically expressed using the parameter association equation as follows:

[0186]

[0187] Among them, p i is the base station coordinate of the i-th base station, p is the center of mass position of the inspection drone, l i is the distance from the center of mass of the inspection drone to the i-th base station, n a is the acceleration measurement noise, n w is the angular velocity measurement noise, for b a The derivative of for b w The derivative of .

[0188] Furthermore, based on the parameter association equation, the state estimation equation and the observation equation are combined to obtain the real-time center of mass position, velocity and attitude of the inspection UAV.

[0189] It can be understood that the parameters in the observation equation are constants or measurable data, and the distance l from the inspection drone to the i-th base station can be obtained according to the observation equation: i , combined with the parameter association equation l i and base station coordinates p i , the relationship between the center of mass position p of the inspection drone and other relationships in the parameter association equation, i Substituting into the state estimation equation, we can get a set of three-variable equations, which can eventually solve the real-time center of mass position, speed and attitude of the inspection drone, that is, the position and attitude data.

[0190] The position and posture data obtained in the above process is the state vector of the filter at time t-1. The state vector of the filter at time t-1 is predicted based on the state vector of the filter at time t, providing more accurate data for adjusting the inspection path, which can further improve the accuracy of the inspection task completion and reduce the possibility of collision of the inspection drone.

[0191] Using the above observation equation and combining it with EKF (Extended Kalman Filter), the linearized equation of motion is given:

[0192] x t|t-1 =Jx t-1 +w,w∈N(0,Q noise )

[0193]

[0194] Among them, x t-1 is the state vector of the filter at time t-1, p t-1 is the center of mass position of the UAV at time t-1, v t-1 is the speed of the drone at time t-1, q t-1 is the attitude of the drone at time t-1, is the distance from the center of mass of the drone to the i-th base station at time t-1, is the zero bias parameter of the accelerometer in the IMU at time t-1, is the zero bias parameter of the gyroscope in the IMU at time t-1, x t|t-1 is the state vector of the filter at time t predicted based on the state vector of the filter at time t-1, p t is the center of mass position of the UAV at time t, v t is the speed of the drone at time t, q t is the attitude of the drone at time t, is the distance from the center of mass of the drone to the i-th base station at time t, is the zero bias parameter of the accelerometer in the IMU at time t, is the zero bias parameter of the gyroscope in the IMU at time t, w is the noise vector of the fusion positioning system and w has a mean of 0 and a covariance of Q noise is the Gaussian distribution of , and J is the linearized Jacobian matrix.

[0195]

[0196] Among them, J is the linearized Jacobian matrix, I is the unit matrix, Δts is the time step, (ab a ) is the measured acceleration value after correction by the zero bias parameter ba, (-ω+b w ) is the zero bias parameter b w

[0197] The corrected angular velocity measurement, Exp() is the natural exponential function, and R is the rotation matrix.

[0198] Based on the linearized equation of motion, the prediction process of the extended Kalman filter is given:

[0199] X pred =x t|t-1

[0200] Among them, x pred It is the state vector for predicting the current state of the inspection drone.

[0201] P pred =P t|t-1 =JP t-1 J T +Q noise

[0202] Among them, P pred is the covariance matrix for predicting the current state of the inspection drone, P t|t-1 is the covariance matrix at time t predicted based on the covariance matrix at time t-1, P t-1 is the covariance matrix updated at time t-1, Q noise is the noise covariance matrix in the prediction process.

[0203] K t =P pred H T (HP pred H T +V) -1

[0204] Among them, K t is the Kalman gain in the prediction process, P pred

[0205] is the covariance matrix for predicting the current state of the inspection UAV, H is the observation matrix, and V is the measurement noise covariance matrix.

[0206] By using the Kalman gain to update the state vector of the inspection drone, the state vector of the inspection drone at the current time, i.e., time t, can be obtained:

[0207] x t =x pred +K t (z t -h(x pred ))

[0208] Among them, x t is the state vector of the inspection drone at the current time, i.e., time t, x pred is the actual measurement value of the UAV, z t is the state vector for predicting the current state of the inspection drone, h(x pred ) is the observed value of the state vector at the current moment.

[0209] After obtaining the state vector of the inspection drone at the current moment, the covariance matrix can be updated and the Kalman gain can be updated to predict the state vector of the inspection drone at the next moment.

[0210] The updated covariance matrix is ​​specifically expressed as:

[0211] P t =(IK t H)P pred

[0212] Among them, P t is the updated value of the covariance matrix at the current time t, and I is the identity matrix.

[0213] S104: Adjust the inspection path of the inspection drone according to the position and posture data to complete the inspection task of the inspection drone.

[0214] Specifically, the obstacle position data corresponding to the position and posture data is obtained, and based on the obstacle position data, a path planning method is used to determine the inspection path.

[0215] Through the microphone array and IMU fusion positioning system, the position and posture data of the inspection drone and the relative position of the obstacle, i.e., the obstacle position data, can be obtained, and the inspection path can be obtained by using a path planning algorithm. In this embodiment, the above-mentioned path planning algorithm can be an A* algorithm or a D* algorithm. In other implementations, the path planning algorithm can also use a simulated annealing algorithm, an artificial potential field method, a fuzzy logic algorithm, a taboo search algorithm, etc., without limitation. During the inspection flight of the inspection drone, the position and posture data and obstacle position data of the inspection drone are detected and updated in real time. When a new obstacle or environmental change is detected, the inspection path is dynamically updated according to the path planning method to ensure that the inspection drone can avoid obstacles safely.

[0216] It is understandable that the current inspection paths of inspection drones are mostly obtained based on static algorithms. When the RTK signal fails, the position of the inspection drone may change and is no longer within the original inspection path. At this time, the inspection path needs to be adjusted according to the position of the inspection drone. The position of the inspection drone can be adjusted back to the original inspection path to continue inspection, or obstacles can be bypassed based on the position of the inspection drone while generating a new inspection path. It is also possible to only adjust the inspection path so that the inspection drone avoids obstacles. There is no limitation on this.

[0217] At the same time, the above positioning status and electromagnetic interference status are monitored in real time. When the above positioning status and electromagnetic interference status are both normal, the inspection drone will switch to RTK navigation; when any of the above positioning status and electromagnetic interference status is abnormal, maintain or switch to the backup navigation system based on the fusion of voiceprint signals and IMU.

[0218] The control method of the above-mentioned inspection drone also includes:

[0219] According to the position and attitude data, a first distance between the position and attitude data and the no-fly zone is given. Based on a preset safety range, the first distance is judged to obtain a first judgment result. According to the first judgment result, an alarm instruction is output.

[0220] Obtain the position coordinates (x a ,y a ,z a ) to determine whether the inspection drone has entered the no-fly zone. The no-fly zone for inspection drones is as follows:

[0221]

[0222]

[0223] Among them, D x is the distance of the first distance in the x direction, D yis the distance of the first distance in the y direction, D z is the distance of the first distance in the z direction.

[0224] In this example, the alarm level is set according to the distance between the inspection UAV position and the no-fly zone, and the settings are as follows:

[0225]

[0226] When the inspection UAV is far from the no-fly zone, the system alarm will not be triggered. When the inspection UAV approaches or enters the no-fly zone, the specific alarm levels are set as follows:

[0227] When the first distance is greater than 5 meters and less than 10 meters, it is set as a level-three alarm; when the first distance is greater than 1 meter and less than 5 meters, it is set as a level-two alarm; when the first distance is less than 1 meter, it is set as a level-one alarm. Different alarm methods are selected according to different alarm levels, and real-time warnings can be sent to the user through the above-mentioned sound alarms, visual alarms, remote notifications, etc., which are not limited here. In a specific example, when the alarm level is a level-three alarm, remote notification is used for alarm; when the alarm level is a level-two alarm, visual alarm is used for alarm; when the alarm level is a level-one alarm, sound alarm is used for alarm. In another specific example, when the alarm level is a level-three alarm, remote notification is used for alarm; when the alarm level is a level-two alarm, both visual alarm and remote notification are used for alarm; when the alarm level is a level-one alarm, both sound alarm, visual alarm and remote notification are used for alarm.

[0228] The above control method of the inspection UAV further includes: giving a second distance between the position coordinates in the position and attitude data and the obstacle position data according to the position and attitude data. Based on a preset obstacle safety range, the second distance is judged to obtain a second judgment result. According to the second judgment result, an alarm instruction is output.

[0229] Obtain the position coordinates and obstacle position data in the position and attitude data of the inspection UAV in real time, and judge whether the distance between the inspection UAV and the obstacle is relatively close.

[0230] Detect the second distance D in real time min , and compare it with the set thresholds Dis1 and Dis2, 0 ≤ Dis1 < Dis2. When D min ≥ Dis2, it means that the inspection UAV maintains a safe distance from the obstacle and no warning is required; when Dis1 ≤ D min < Dis2, it means that the distance between the inspection UAV and the obstacle is not completely safe, and a level-two warning is issued; when D minWhen <Dis1>, it indicates that the inspection UAV is too close to the equipment or obstacle, and a first-level warning is issued. The above obstacle safety range is determined by the thresholds <Dis1> and <Dis2>, and the thresholds <Dis1> and <Dis2> can be set according to factors such as the actual scenario and the inspection speed of the inspection UAV, and no limitation is made thereto.

[0231] Refer to Figure 2 , a control method for an inspection UAV provided by the present invention obtains real-time positioning signal data and electromagnetic signal data through various sensors configured on the inspection UAV, respectively analyzes the positioning signal data and the electromagnetic signal data, obtains a plurality of signal features and electromagnetic features and performs weighted fusion respectively, and judges the real-time positioning inspection state of the inspection UAV based on the result of the weighted fusion, so as to timely understand the influence of various factors in the environment on the inspection UAV during the inspection process, judge whether there are problems of electromagnetic interference and / or RTK failure of the inspection UAV, and based on the problems existing in the inspection of the inspection UAV, start the fusion positioning, combine the voiceprint signal and the IMU to determine the position and attitude data of the inspection UAV, and timely adjust the inspection path to ensure the accurate completion of the inspection task.

[0232] Refer to Figure 3 , an embodiment of the present invention provides a control device for an inspection UAV, including:

[0233] A data acquisition module 201, configured to acquire inspection sensing data, where the inspection sensing data is data collected by sensors during the inspection process of the inspection UAV;

[0234] A state determination module 202, configured to analyze the inspection sensing data and determine the real-time positioning inspection state of the inspection UAV;

[0235] A position analysis module 203, configured to extract the voiceprint signal of the inspection UAV based on the real-time positioning inspection state of the inspection UAV, and combine the state estimation equation of the inertial measurement unit to determine the position and attitude data of the inspection UAV;

[0236] A path adjustment module 204, configured to adjust the inspection path of the inspection UAV according to the position and attitude data to complete the inspection task of the inspection UAV.

[0237] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the described modules can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0238] Although preferred embodiments of the present invention have been described, additional changes and modifications may be made to these embodiments by those skilled in the art once the basic inventive concepts are known. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention. Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A control method for an inspection drone, characterized in that: include: Acquire patrol sensor data, where the patrol sensor data is data collected by sensors of the patrol drone during the patrol process; Analyze inspection sensor data to determine the real-time positioning and inspection status of the inspection drone; Based on the real-time positioning and inspection status of the inspection drone, the voiceprint signal of the inspection drone is extracted, and combined with the state estimation equation of the inertial measurement unit, the position and attitude data of the inspection drone is determined; According to the position and posture data, the inspection path of the inspection drone is adjusted to complete the inspection mission of the inspection drone.

2. The control method of the inspection drone according to claim 1, characterized in that: Inspection sensor data includes positioning signal data and electromagnetic signal data; The analyzing the inspection sensor data to determine the real-time positioning inspection status of the inspection drone includes: Extract signal features from positioning signal data and give positioning error; Extract electromagnetic features from electromagnetic signal data and give electromagnetic interference intensity; The positioning error and the electromagnetic interference intensity are judged respectively to obtain the positioning state and the electromagnetic interference state; Based on the positioning status and / or electromagnetic interference status, the real-time positioning and inspection status of the inspection drone is determined.

3. The control method of the inspection drone according to claim 2, characterized in that: The positioning signal data includes satellite signal power, and the signal characteristics include signal strength characteristics, multipath characteristics, satellite spatial distribution characteristics, carrier phase error and delay characteristics; The extracting signal features from the positioning signal data and providing the positioning error comprises: Obtain satellite spatial distribution characteristics, carrier phase error and delay characteristics; According to the satellite signal power, the ideal attenuation value, actual attenuation value and additional attenuation value of the satellite signal are given; Determine the signal strength characteristics of the satellite signal based on the ideal attenuation value, the actual attenuation value and the additional attenuation value of the satellite signal; Analyze signal strength characteristics and give multipath characteristics; Based on the weight of each signal feature, the signal strength feature, multipath feature, satellite spatial distribution feature, carrier phase error and delay feature are integrated to obtain the positioning error.

4. The control method of the inspection drone according to claim 3, characterized in that: The satellite spatial distribution characteristics are indexed by the geometrical dilution of precision; Set the judgment threshold of geometric precision dilution; Based on the analysis of geometric precision factor and discrimination threshold, the influence of geometric precision factor on positioning error is measured; The geometric precision factor is specifically expressed as: Where GDOP is the geometric dilution of precision, q 11 is the error of the receiver’s position coordinates in the x-axis direction, q 22 is the error of the receiver’s position coordinates in the y-axis direction, q 33 is the error of the receiver’s position coordinates in the z-axis direction, q 44 is the time error of the receiver.

5. The control method of the inspection drone according to claim 2, characterized in that: The electromagnetic characteristics include magnetic field intensity characteristics, magnetic field spectrum characteristics and magnetic field change rate characteristics; The step of extracting electromagnetic features from electromagnetic signal data and providing electromagnetic interference intensity comprises: Based on the electromagnetic signal data, the magnetic field intensity characteristics, magnetic field spectrum characteristics and magnetic field change rate characteristics are given, wherein the magnetic field spectrum characteristics are obtained by performing fast Fourier transform on the electromagnetic field signal, and the magnetic field change rate characteristics are the rate of change of the electromagnetic field intensity over time; Based on the weights corresponding to each electromagnetic feature, the magnetic field intensity feature, magnetic field spectrum feature and magnetic field change rate feature are integrated to obtain the electromagnetic interference intensity.

6. The control method of the inspection drone according to claim 1, characterized in that: The real-time positioning inspection status of the inspection drone is extracted based on the inspection drone's voiceprint signal, and the position and posture data of the inspection drone is determined in combination with the state estimation equation of the inertial measurement unit, including: When the real-time positioning inspection status is abnormal, the acoustic signal features in the voiceprint signal are extracted and the observation equation is established; Combined with the state estimation equation of the inertial measurement unit, the real-time position and real-time attitude of the inspection UAV are analyzed to obtain the position and attitude data of the inspection UAV.

7. The control method of the inspection drone according to claim 6, characterized in that: The state estimation equation of the inertial measurement unit is combined to analyze the real-time position and real-time attitude of the inspection drone to obtain the position and attitude data of the inspection drone, including: Obtaining a parameter association equation, wherein the parameter association equation includes an association relationship between each parameter in the observation equation and each parameter in the state estimation equation; Based on the parameter association equation, the observation equation and the state estimation equation are integrated to analyze the real-time position and attitude of the inspection UAV and determine the position and attitude data of the inspection UAV.

8. The control method of the inspection drone according to claim 1, characterized in that: The method further comprises: According to the position and attitude data, a first distance between the position and attitude data and a no-fly zone is given; Based on the preset safety range, the first distance is judged to obtain a first judgment result; According to the first judgment result, an alarm instruction is output.

9. The control method of the inspection drone according to claim 1, characterized in that: The step of adjusting the inspection path of the inspection drone according to the position and posture data to complete the inspection task of the inspection drone includes: Obtain obstacle position data corresponding to the position and posture data; Adjust the inspection path based on obstacle location data.

10. A control device for an inspection drone, characterized in that: include: A data acquisition module is used to acquire patrol sensor data, wherein the patrol sensor data is data collected by sensors of the patrol drone during the patrol process; The status determination module is used to analyze the inspection sensor data and determine the real-time positioning inspection status of the inspection drone; The position analysis module is used to extract the voiceprint signal of the inspection drone based on the real-time positioning inspection status of the inspection drone, and determine the position and attitude data of the inspection drone in combination with the state estimation equation of the inertial measurement unit; The path adjustment module is used to adjust the inspection path of the inspection drone according to the position and posture data to complete the inspection task of the inspection drone.

Images