Target equipment detection control method and device

Abstract

The invention provides a target equipment detection control method and device, and relates to the technical field of control. The method comprises the steps that in the cruising process of an unmanned aerial vehicle carrying a mechanical arm, the motion state quantity of the unmanned aerial vehicle is controlled through a PID controller, so that the unmanned aerial vehicle travels to a hovering operation position; wherein the motion state quantity comprises a flight attitude control quantity and a flight height control quantity; and controlling the rotation angular velocity and three-dimensional pose information of a rotor wing of the unmanned aerial vehicle at the hovering operation position through a neural network model so as to enable the unmanned aerial vehicle to hover at the hovering operation position, and controlling the rotation angle of the mechanical arm to rotate so as to detect target equipment through the mechanical arm. The device executes the method. According to the target equipment detection control method and device provided by the embodiment of the invention, the target equipment can be safely and efficiently detected and controlled.

Description

Technical Field

[0001] The present invention relates to the field of control technology, and in particular to a target device detection control method and device. Background Art

[0002] In the complex and hazardous operation scenarios of the power industry, new robot technology has huge application potential and demand. The maintenance technology, safety assurance and intelligent requirements of power equipment are gradually improving, and the application of new robot technology in this field will be more extensive and in-depth. However, this field often faces harsh operating environments and multiple potential dangerous factors. Equipment maintenance and overhaul work requires highly specialized personnel to carry out in a high-risk environment, which poses a series of technical and operational challenges such as personal safety hazards and low efficiency.

[0003] At present, rotary-wing drones have been widely used in many fields, especially in monitoring and perception. The various sensors they carry play a role in data collection and environmental status perception. However, these applications mainly stay on the passive collection of environmental information and cannot achieve direct interaction with the external environment. Drones equipped with robotic arms have shown greater potential and flexibility. Robotic arm drones can not only perform aerial monitoring, but also directly grasp and move target objects, and perform various complex tasks in areas that are difficult for people to reach. For this reason, drones with robotic arms have received more and more attention and application in recent years. However, the operating environment of the power industry is complex, and there are great difficulties in how to achieve safe and efficient equipment inspection by controlling drones with robotic arms. Summary of the invention

[0004] In view of the problems in the prior art, an embodiment of the present invention provides a target device detection control method and apparatus, which can at least partially solve the problems in the prior art.

[0005] In one aspect, the present invention provides a target device detection control method, comprising:

[0006] During the cruising process of the UAV equipped with the mechanical arm, the motion state quantity of the UAV is controlled by a PID controller so that the UAV travels to a hovering operation position;

[0007] Wherein, the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity;

[0008] At the hovering operation position, the rotor rotation angular velocity and three-dimensional posture information of the UAV are controlled by a neural network model so that the UAV hovers at the hovering operation position, and the rotation angle of the robotic arm is controlled to perform a rotation action so that the target device can be detected by the robotic arm.

[0009] The neural network model is obtained in advance by fusing the output results corresponding to a plurality of RBF models; each RBF model is obtained by training a pre-assigned sample subset.

[0010] The neural network model is obtained by fusing the output results corresponding to multiple RBF models, including:

[0011] The neural network model is obtained by fusing the output results corresponding to multiple RBF models respectively through a linear weighted sum method.

[0012] Wherein, before obtaining each RBF model through pre-allocated sample subset training, the target device detection control method further includes:

[0013] Extracting production training samples corresponding to each application scenario from the sample set; the application scenarios include cruising scenarios and hovering operation scenarios;

[0014] In each production training sample, samples are randomly selected from the corresponding production training samples according to the preset sample subset size and the preset sample subset number to form sample subsets corresponding to each production training sample and the number of samples in each sample subset is equal to the preset sample subset size.

[0015] The step of randomly selecting samples from the corresponding production training samples according to the preset sample subset size and the preset sample subset number in each production training sample includes:

[0016] In each production training sample, samples are randomly selected with replacement from the corresponding production training sample according to the preset sample subset size and the preset sample subset number.

[0017] Wherein, the PID controller is a fuzzy controller that performs fuzzy reasoning through a fuzzy rule table.

[0018] In one aspect, the present invention provides a target device detection control device, comprising:

[0019] A first control unit is used to control the motion state of the UAV equipped with the mechanical arm through a PID controller during the cruising process of the UAV so as to make the UAV travel to a hovering operation position;

[0020] Wherein, the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity;

[0021] The second control unit is used to control the rotor rotation angular velocity and three-dimensional posture information of the UAV through a neural network model at the hovering operation position so that the UAV hovers at the hovering operation position, and to control the rotation angle of the robotic arm to perform a rotation action so that the target device can be detected through the robotic arm.

[0022] In another aspect, an embodiment of the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following method when executing the computer program:

[0023] During the cruising process of the UAV equipped with the mechanical arm, the motion state quantity of the UAV is controlled by a PID controller so that the UAV travels to a hovering operation position;

[0024] Wherein, the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity;

[0025] At the hovering operation position, the rotor rotation angular velocity and three-dimensional posture information of the UAV are controlled by a neural network model so that the UAV hovers at the hovering operation position, and the rotation angle of the robotic arm is controlled to perform a rotation action so that the target device can be detected by the robotic arm.

[0026] An embodiment of the present invention provides a computer-readable storage medium, including:

[0027] The computer readable storage medium stores a computer program, and when the computer program is executed by a processor, the following method is implemented:

[0028] During the cruising process of the UAV equipped with the mechanical arm, the motion state quantity of the UAV is controlled by a PID controller so that the UAV travels to a hovering operation position;

[0029] Wherein, the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity;

[0030] At the hovering operation position, the rotor rotation angular velocity and three-dimensional posture information of the UAV are controlled by a neural network model so that the UAV hovers at the hovering operation position, and the rotation angle of the robotic arm is controlled to perform a rotation action so that the target device can be detected by the robotic arm.

[0031] The embodiment of the present invention further provides a computer program product, the computer program product comprising a computer program, and when the computer program is executed by a processor, the following method is implemented:

[0032] During the cruising process of the UAV equipped with the mechanical arm, the motion state quantity of the UAV is controlled by a PID controller so that the UAV travels to a hovering operation position;

[0033] Wherein, the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity;

[0034] At the hovering operation position, the rotor rotation angular velocity and three-dimensional posture information of the UAV are controlled by a neural network model so that the UAV hovers at the hovering operation position, and the rotation angle of the robotic arm is controlled to perform a rotation action so that the target device can be detected by the robotic arm.

[0035] The target device detection control method and device provided by the embodiment of the present invention control the motion state quantity of the unmanned aerial vehicle equipped with a mechanical arm through a PID controller during the cruising process of the unmanned aerial vehicle equipped with a mechanical arm, so that the unmanned aerial vehicle travels to a hovering operation position; wherein the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity; at the hovering operation position, the rotor rotation angular velocity and three-dimensional posture information of the unmanned aerial vehicle are controlled through a neural network model, so that the unmanned aerial vehicle hovers at the hovering operation position, and the rotation angle of the mechanical arm is controlled to perform a rotational motion, so that the target device can be detected by the mechanical arm, and the target device detection control can be performed safely and efficiently. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0037] Figure 1 It is a flowchart of a target device detection control method provided by an embodiment of the present invention.

[0038] Figure 2 It is a flowchart of a target device detection control method provided by another embodiment of the present invention.

[0039] Figure 3 It is a flowchart of a target device detection control method provided by another embodiment of the present invention.

[0040] Figure 4 It is a structural schematic diagram of a target device detection control device provided by an embodiment of the present invention.

[0041] Figure 5 A schematic diagram of the physical structure of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical scheme and advantages of the embodiments of the present invention more clear, the embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention. It should be noted that, in the absence of conflict, the embodiments in this application and the features in the embodiments can be combined with each other arbitrarily.

[0043] Explanation of related terms:

[0044] Robotic arm drone: refers to a drone equipped with a robotic arm. In addition to conventional flight capabilities, drones have expanded their operational functions and mission execution range by integrating robotic arm technology. Drones with arms can perform more complex tasks in the air, such as grabbing and manipulating objects. Drones with robotic arms are more complex to design and control because they need to manage flight stability and precise operation of the robotic arm at the same time to ensure that the drone remains stable during operation. This type of drone also needs to be equipped with advanced sensors and computing systems for precise navigation, positioning and robotic arm control.

[0045] Neural network controller: This controller is a control system based on neural network. By training the neural network model, the input data is mapped to the output control signal, thereby achieving control of complex systems.

[0046] PID control: PID is an automatic control method, where "PID" stands for proportional, integral and differential. The control method controls the controlled object by linearly combining the control deviation between the given value and the actual output value into a control quantity in proportion, integral and differential.

[0047] Figure 1 FIG. 1 is a flow chart of a target device detection control method provided by an embodiment of the present invention. Figure 1 As shown, the target device detection control method provided by the embodiment of the present invention includes:

[0048] Step S1: During the cruising process of the UAV equipped with the robotic arm, the motion state of the UAV is controlled by a PID controller so that the UAV travels to a hovering operation position;

[0049] The motion state quantity includes a flight attitude control quantity and a flight altitude control quantity.

[0050] Step S2: At the hovering operation position, the rotor rotation angular velocity and three-dimensional posture information of the UAV are controlled by a neural network model so that the UAV hovers at the hovering operation position, and the rotation angle of the robotic arm is controlled to perform a rotation motion so that the target device can be detected by the robotic arm.

[0051] In the above step S1, the device controls the motion state of the drone equipped with the robotic arm through a PID controller during the cruising process of the drone, so that the drone travels to a hovering operation position;

[0052] The motion state quantity includes the flight attitude control quantity and the flight altitude control quantity. The device may be a computer device for executing the method, such as a server. The acquisition, storage, use, and processing of data in the technical solution of this application are in compliance with relevant regulations.

[0053] In the above step S2, the device controls the rotor rotation angular velocity and three-dimensional posture information of the drone at the hovering operation position through a neural network model, so that the drone hovers at the hovering operation position, and controls the rotation angle of the mechanical arm to perform a rotation action, so that the target device can be detected by the mechanical arm. Figure 2 As shown, the method of the present invention combines the neural network model and the PID controller to receive the UAV state feedback and the robotic arm state feedback to control the UAV and the robotic arm.

[0054] The kinetic model is described as follows:

[0055] The structure of the drone and robotic arm is designed according to the operation requirements of the flying machine in the GIS equipment area of ​​the substation. Although the modeling process of the three-degree-of-freedom robotic arm is more complicated, it is more conducive to participating in tasks such as manipulation and assembly when mounted on the drone, and has higher flexibility. The front end mechanism of the drone equipped with the three-degree-of-freedom robotic arm is equipped with an ultrasonic sensor.

[0056] The drone and the robot arm are modeled and controlled as a whole. First, the rotation matrix R is defined:

[0057]

[0058] Define the control matrix:

[0059]

[0060] Where: p is the position of the rotorcraft, Φ is the Euler angle of the rotorcraft in the inertial coordinate system, and θ is the rotation angle of the three robotic arms. The speed of the rotorcraft can be expressed as:

[0061]

[0062] The translational velocity of the center of mass is described by the inertial coordinate system and the rotor body coordinate system respectively. ω, ω B are the angular velocities in the inertial coordinate system and the rotor body coordinate system respectively.

[0063] Finally, using the form containing the variable q express:

[0064]

[0065] For each robot arm, the translation velocity and angular velocity are respectively obtained by the joint variables and the corresponding Jacobian matrix J 1j ,J rj The overall kinematic model can be obtained by association as follows:

[0066]

[0067] The neural network model is obtained in advance by fusing the output results corresponding to a plurality of RBF models; each RBF model is obtained by training a pre-assigned sample subset.

[0068] The neural network model is obtained by fusing the output results corresponding to multiple RBF models, including:

[0069] The neural network model is obtained by fusing the output results corresponding to multiple RBF models respectively through a linear weighted sum method.

[0070] After the given expected command signal, the UAV with a robotic arm calculates the angle (rotation angle of the robotic arm), speed (rotation angular velocity of the rotor) and posture (three-dimensional posture information) through designing and building a neural network controller to complete the GIS equipment inspection work in the substation. The integrated RBF neural network structure is constructed based on the idea of ​​ensemble learning to improve the accuracy of control law approximation.

[0071] RBF network belongs to multi-layer feedforward network, which usually includes three layers, namely input layer, hidden layer and output layer. The function of input layer is mainly to accept data input. Each unit of hidden layer is called hidden unit, which provides a radial basis function. Generally, Gaussian function is selected as the basis function of hidden layer. The function of hidden layer is to realize the transformation from input space to hidden space. The mapping between hidden space and output space is linear mapping, that is, the output of network is the linear weighted sum of hidden units.

[0072] like Figure 3 As shown, the output result obtains the combination method of multiple RBF networks. Assuming that there are m RBF networks to be combined, the combination method is to calculate the linear weighted sum of the m RBF networks.

[0073]

[0074] y i (i=1,2,…,n) are the predicted outputs of m RBF networks, w i for y i The weight of It is the combined output of multiple RBF networks. The key to combining multiple RBF networks is to determine the weight w i The method is the averaging ensemble method (SAE for short).

[0075] w 1 =w 2 =…=w m =1 / m

[0076]

[0077] Before obtaining each RBF model through pre-assigned sample subset training, the target device detection control method further includes:

[0078] Extracting production training samples corresponding to each application scenario from the sample set; the application scenarios include cruising scenarios and hovering operation scenarios;

[0079] In each production training sample, samples are randomly selected from the corresponding production training samples according to the preset sample subset size and the preset sample subset number to form sample subsets corresponding to each production training sample and the number of samples in each sample subset is equal to the preset sample subset size.

[0080] The randomly selecting samples from the corresponding production training samples according to the preset sample subset size and the preset sample subset number in each production training sample includes:

[0081] In each production training sample, samples are randomly selected with replacement from the corresponding production training sample according to the preset sample subset size and the preset sample subset number.

[0082] The cruising scenario corresponds to controlling the motion state of the UAV through a PID controller; the hovering operation scenario corresponds to controlling the rotor rotation angular velocity and three-dimensional posture information of the UAV through a neural network model, as well as controlling the rotation angle of the robotic arm to perform rotational movements.

[0083] like Figure 3 As shown, the preset sample subset size corresponds to the number of samples in each sample, and the number of samples in sample 1 to sample n can be set to be the same. The preset sample subset number corresponds to n.

[0084] (1) Set the preset sample subset size and the preset sample subset number.

[0085] (2) Randomly select (with replacement) a sample point from the production training sample and assign it to a sample.

[0086] (3) Repeat step (2) until each sample is fully allocated.

[0087] The PID controller is a fuzzy controller that performs fuzzy reasoning through a fuzzy rule table.

[0088] Constructing UAV fuzzy PID controller When the PID control algorithm is applied in a closed-loop system, the input signal it receives is the error e between the expected value and the output value of the controlled object, and then the control law of the system is obtained through proportional control, integral control and differential control to achieve the control target. The fuzzy PID controller uses fuzzy rules to self-tune its control parameters on the basis of PID control so that the system can reach the optimal solution more quickly.

[0089] The error e and the error change rate ec between the set system expected value and the system output value are then set, and fuzzy reasoning is performed using the pre-set fuzzy control rule table to obtain the three parameter increments Δkp, Δki, and Δkd of the PID controller. Finally, these three increments are integrated into the PID controller to obtain a PID control system with self-tuning parameters.

[0090] The PID controller consists of a proportional link (P), an integral link (I), and a differential link (D). The control law of the controlled object after being solved by the PID control algorithm can be expressed as:

[0091]

[0092] In the formula, K p is the proportionality coefficient, T i is the integration time constant, T d is the differential time constant, and e(t) is the error between the system expected value and the system output value.

[0093] The control steps of the fuzzy controller are as follows: first, the input and output variables of the controller need to be determined, and then the clear input quantity is fuzzified by the determined membership function, and then the fuzzy output quantity is obtained by fuzzy reasoning through the fuzzy rule table, and finally the output fuzzy quantity is defuzzified, converted into a clear quantity and then transmitted to the controlled object, thereby controlling the controlled object to achieve the desired goal.

[0094] PID controllers are used for attitude control and altitude control. By adjusting the proportional, integral and differential parameters, precise control of the UAV's flight attitude can be achieved. In the substation GIS environmental inspection, the parameters of the PID controller used in the robot arm UAV flight are adjusted according to the specific characteristics of the UAV and the flight environment.

[0095] When the system encounters large disturbances or significant nonlinear characteristics, such as the hovering operation stage, it switches to neural network control and uses the nonlinear mapping capabilities of the neural network model to handle such conditions.

[0096] In the hovering operation of the UAV, the neural network control model can better adapt to environmental changes and external disturbances, and improve the stability and accuracy of the hovering. When the robotic arm UAV reaches the inspection point and switches to the hovering mode, the controller uses a neural network controller. By training and optimizing the neural network controller using the ensemble learning idea, the optimal control strategy for the hovering of the UAV can be learned. The system convergence speed is improved, the hovering error can be reduced, and the corresponding performance index requirements of the UAV hovering operation can be achieved.

[0097] After receiving the given expected command signal, the UAV with a robotic arm will use specially designed conventional PID controllers and neural network controllers to accurately calculate the position and attitude control of the rotor UAV and the control of the robotic arm. Subsequently, the torque distributor will accurately control the speed of the brushless DC motor of the rotor UAV and the rotation angle of the robotic arm servo. These control instructions will be applied to the mathematical model of the UAV with a robotic arm to achieve complex tasks such as flight and ultrasonic detection of the robotic arm. After the task is completed, the system will provide comprehensive feedback to the control end based on the actual output results and combined with external environmental disturbance factors, thereby forming a closed-loop control system to ensure the accuracy of the control results and the stability of the system.

[0098] The target device detection control method provided by the embodiment of the present invention can solve the following existing technical problems:

[0099] 1. Accuracy issue: UAV inspection requires a high-precision positioning and navigation system to ensure accurate flight control and positioning technology research on the flying robot in the GIS equipment area of ​​the substation. The core control system and station base station auxiliary positioning technology are used to achieve accurate flight and positioning in the substation to ensure a safe distance.

[0100] 2. Interference with robot arm operation: When and after the robot arm comes into contact with the environment, the sudden changes or disturbances in the forces and torques and the center of gravity of the system require the drone to continue to hover and resist aerodynamic effects to achieve stable grasping of the target.

[0101] 3. Complex coordination and control: The motion mechanism of the rotorcraft itself has strong nonlinear and strong coupling characteristics. When adding an object to be manipulated by the actuator arm, this coupling effect will be further aggravated, bringing difficulties to both modeling and control.

[0102] The target device detection control method provided by the embodiment of the present invention combines the original PID control with the neural network controller, and utilizes the stronger adaptive performance of the neural network controller. The information required for the flight of the UAV is complex and changeable, and it changes continuously with the changes in the environment. The deep neural network can adapt to new environmental changes through training, thereby improving the adaptive ability of the navigation control system, enabling it to better adapt to different environmental conditions. This improvement in adaptive performance helps the UAV maintain a stable flight state in various complex environments.

[0103] The target device detection control method provided by the embodiment of the present invention has the following beneficial technical effects:

[0104] 1. Use a neural network controller in the robotic arm unmanned flying and hovering operation scenario to achieve arbitrary nonlinear mapping. This gives the neural network controller a significant advantage in dealing with complex nonlinear systems and can more accurately simulate and control the dynamic behavior of the actual system.

[0105] 2. In the face of system uncertainty and interference, neural network control can effectively improve the accuracy of the control model through data accumulation, ensuring the stability and robustness of the control effect. This robustness makes the neural network controller more reliable and stable in practical applications.

[0106] 3. Improve system stability. When facing noise, interference or parameter changes, the neural network can enhance robustness by learning the characteristics of the data set and adaptively deal with the uncertainty of the system. The PID controller can provide precise control in a stable environment, reduce overshoot and oscillation, and maintain system stability.

[0107] 4. Selecting the appropriate controller according to different working conditions can ensure that the system always operates under the optimal control strategy, thereby improving the overall control performance. When dealing with complex nonlinear problems, its parallel processing capability and nonlinear mapping capability can significantly improve the efficiency of problem solving. The PID control algorithm is simple and has low computational complexity, making it suitable for scenarios that require fast response and real-time control.

[0108] The target device detection control method provided by the embodiment of the present invention controls the motion state quantity of the unmanned aerial vehicle equipped with a mechanical arm through a PID controller during the cruising process of the unmanned aerial vehicle equipped with a mechanical arm, so that the unmanned aerial vehicle travels to a hovering operation position; wherein the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity; at the hovering operation position, the rotor rotation angular velocity and three-dimensional posture information of the unmanned aerial vehicle are controlled through a neural network model, so that the unmanned aerial vehicle hovers at the hovering operation position, and the rotation angle of the mechanical arm is controlled to perform a rotational motion, so that the target device can be detected by the mechanical arm, and the target device detection control can be performed safely and efficiently.

[0109] Furthermore, the neural network model is obtained in advance by fusing the output results corresponding to a plurality of RBF models respectively; each RBF model is obtained by training a pre-assigned sample subset respectively. The above-mentioned embodiment can be referred to for explanation and will not be described in detail.

[0110] Furthermore, the neural network model is obtained by fusing the output results corresponding to the multiple RBF models, including:

[0111] The neural network model is obtained by fusing the output results corresponding to the multiple RBF models respectively through the linear weighted sum method. The above-mentioned embodiment can be referred to for explanation and will not be described in detail.

[0112] Furthermore, before obtaining each RBF model through pre-assigned sample subset training, the target device detection control method further includes:

[0113] Production training samples corresponding to each application scenario are extracted from the sample set; the application scenarios include cruising scenarios and hovering operation scenarios; the description can be made with reference to the above embodiment and will not be repeated here.

[0114] In each production training sample, samples are randomly selected from the corresponding production training samples according to the preset sample subset size and the preset sample subset number, forming sample subsets corresponding to each production training sample in number equal to the preset sample subset number, and the number of samples in each sample subset is equal to the preset sample subset size. The above embodiment can be referred to for explanation and will not be repeated here.

[0115] Furthermore, the randomly selecting samples from the corresponding production training samples according to the preset sample subset size and the preset sample subset number in each production training sample includes:

[0116] In each production training sample, samples are randomly selected with replacement from the corresponding production training samples according to the preset sample subset size and the preset sample subset number. The above description can be referred to and will not be repeated here.

[0117] Furthermore, the PID controller is a fuzzy controller that performs fuzzy reasoning through a fuzzy rule table. The above embodiments can be referred to for explanation and will not be described in detail.

[0118] Figure 4 is a schematic diagram of the structure of a target device detection control device provided by an embodiment of the present invention, such as Figure 4 As shown, the target device detection control device provided by the embodiment of the present invention includes a first control unit 401 and a second control unit 402, wherein:

[0119] The first control unit 401 is used to control the motion state quantity of the UAV equipped with a robotic arm through a PID controller during the cruising process of the UAV, so that the UAV can travel to a hovering operation position; wherein the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity; the second control unit 402 is used to control the rotor rotation angular velocity and three-dimensional posture information of the UAV through a neural network model at the hovering operation position, so that the UAV can hover at the hovering operation position, and control the rotation angle of the robotic arm to perform a rotation action, so that the target device can be detected by the robotic arm.

[0120] Specifically, the first control unit 401 in the device is used to control the motion state quantity of the UAV equipped with a robotic arm through a PID controller during the cruising process of the UAV, so that the UAV can travel to a hovering operation position; wherein the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity; the second control unit 402 is used to control the rotor rotation angular velocity and three-dimensional posture information of the UAV through a neural network model at the hovering operation position, so that the UAV can hover at the hovering operation position, and control the rotation angle of the robotic arm to perform a rotation action, so that the target device can be detected by the robotic arm.

[0121] The target device detection control device provided by the embodiment of the present invention controls the motion state quantity of the unmanned aerial vehicle equipped with a mechanical arm through a PID controller during the cruising process of the unmanned aerial vehicle, so that the unmanned aerial vehicle travels to a hovering operation position; wherein the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity; at the hovering operation position, the rotor rotation angular velocity and three-dimensional posture information of the unmanned aerial vehicle are controlled through a neural network model, so that the unmanned aerial vehicle hovers at the hovering operation position, and the rotation angle of the mechanical arm is controlled to perform a rotational motion, so that the target device can be detected by the mechanical arm, and the target device detection control can be performed safely and efficiently.

[0122] The target device detection control apparatus provided in the embodiment of the present invention can be used to execute the processing flow of the above-mentioned method embodiments. Its functions are not described in detail here, and reference can be made to the detailed description of the above-mentioned method embodiments.

[0123] Figure 5 A schematic diagram of the physical structure of a computer device provided in an embodiment of the present invention is shown in FIG. Figure 5 As shown, the computer device includes: a memory 501, a processor 502, and a computer program stored in the memory 501 and executable on the processor 502. When the processor 502 executes the computer program, the following method is implemented:

[0124] During the cruising process of the UAV equipped with the mechanical arm, the motion state quantity of the UAV is controlled by a PID controller so that the UAV travels to a hovering operation position;

[0125] Wherein, the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity;

[0126] At the hovering operation position, the rotor rotation angular velocity and three-dimensional posture information of the UAV are controlled by a neural network model so that the UAV hovers at the hovering operation position, and the rotation angle of the robotic arm is controlled to perform a rotation action so that the target device can be detected by the robotic arm.

[0127] This embodiment discloses a computer program product, the computer program product including a computer program, and when the computer program is executed by a processor, the following method is implemented:

[0128] During the cruising process of the UAV equipped with the mechanical arm, the motion state quantity of the UAV is controlled by a PID controller so that the UAV travels to a hovering operation position;

[0129] Wherein, the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity;

[0130] At the hovering operation position, the rotor rotation angular velocity and three-dimensional posture information of the UAV are controlled by a neural network model so that the UAV hovers at the hovering operation position, and the rotation angle of the robotic arm is controlled to perform a rotation action so that the target device can be detected by the robotic arm.

[0131] This embodiment provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following method is implemented:

[0132] During the cruising process of the UAV equipped with the mechanical arm, the motion state quantity of the UAV is controlled by a PID controller so that the UAV travels to a hovering operation position;

[0133] Wherein, the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity;

[0134] At the hovering operation position, the rotor rotation angular velocity and three-dimensional posture information of the UAV are controlled by a neural network model so that the UAV hovers at the hovering operation position, and the rotation angle of the robotic arm is controlled to perform a rotation action so that the target device can be detected by the robotic arm.

[0135] Compared with the technical solutions in the prior art, the target device detection control method provided by the embodiment of the present invention controls the motion state quantity of the unmanned aerial vehicle equipped with a mechanical arm through a PID controller during the cruising process of the unmanned aerial vehicle equipped with a mechanical arm, so that the unmanned aerial vehicle travels to a hovering operation position; wherein the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity; at the hovering operation position, the rotor rotation angular velocity and three-dimensional posture information of the unmanned aerial vehicle are controlled through a neural network model, so that the unmanned aerial vehicle hovers at the hovering operation position, and the rotation angle of the mechanical arm is controlled to perform a rotational motion, so that the target device can be detected through the mechanical arm, and the target device detection control can be performed safely and efficiently.

[0136] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0137] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0138] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1A function specified in one or more boxes.

[0139] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0140] In the description of this specification, the description with reference to the terms "one embodiment", "a specific embodiment", "some embodiments", "for example", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0141] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A target device detection control method, characterized in that: include: During the cruising process of the UAV equipped with the mechanical arm, the motion state quantity of the UAV is controlled by a PID controller so that the UAV travels to a hovering operation position; Wherein, the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity; At the hovering operation position, the rotor rotation angular velocity and three-dimensional posture information of the UAV are controlled by a neural network model so that the UAV hovers at the hovering operation position, and the rotation angle of the robotic arm is controlled to perform a rotation action so that the target device can be detected by the robotic arm.

2. The target device detection control method according to claim 1, characterized in that: The neural network model is obtained in advance by fusing the output results corresponding to a plurality of RBF models; each RBF model is obtained by training a pre-assigned sample subset.

3. The target device detection control method according to claim 2, characterized in that: The neural network model is obtained by fusing the output results corresponding to multiple RBF models, including: The neural network model is obtained by fusing the output results corresponding to multiple RBF models respectively through a linear weighted sum method.

4. The target device detection control method according to claim 2, characterized in that: Before obtaining each RBF model through pre-assigned sample subset training, the target device detection control method further includes: Extracting production training samples corresponding to each application scenario from the sample set; the application scenarios include cruising scenarios and hovering operation scenarios; In each production training sample, samples are randomly selected from the corresponding production training samples according to the preset sample subset size and the preset sample subset number to form sample subsets corresponding to each production training sample and the number of samples in each sample subset is equal to the preset sample subset size.

5. The target device detection control method according to claim 4, characterized in that: The randomly selecting samples from the corresponding production training samples according to the preset sample subset size and the preset sample subset number in each production training sample includes: In each production training sample, samples are randomly selected with replacement from the corresponding production training sample according to the preset sample subset size and the preset sample subset number.

6. The target device detection control method according to claim 1, characterized in that: The PID controller is a fuzzy controller that performs fuzzy reasoning through a fuzzy rule table.

7. A target device detection control device, characterized in that: include: A first control unit is used to control the motion state of the UAV equipped with the mechanical arm through a PID controller during the cruising process of the UAV so as to make the UAV travel to a hovering operation position; Wherein, the motion state quantity includes a flight attitude control quantity and a flight altitude control quantity; The second control unit is used to control the rotor rotation angular velocity and three-dimensional posture information of the UAV through a neural network model at the hovering operation position so that the UAV hovers at the hovering operation position, and to control the rotation angle of the robotic arm to perform a rotation action so that the target device can be detected through the robotic arm.

8. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images