VR immersive control system for unmanned aerial vehicle control
By combining VR helmet, experience seat and tactile feedback system in VR system, using an adaptively weighted multi-sensor data fusion method to provide precise tactile feedback, and built-in autonomous obstacle avoidance system in the drone, the haptic feeling in VR system is solved Inadequate feedback and drone control safety issues have achieved a more realistic immersion and safe flight.
Patent Information
- Application Number
- CN202411142013.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-20
- Publication Date
- 2024-11-22
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing VR system lacks an effective tactile feedback mechanism, which weakens the authenticity of the overall immersion. The drone is prone to collision and fall problems when navigating autonomously or controlled by non-professional personnel, which poses safety hazards.
Combining VR helmets, experience seats, drones and haptic feedback systems, an adaptively weighted multi-sensor data fusion method provides precise haptic feedback, and an autonomous obstacle avoidance system is built into the drone, using obstacle avoidance path prediction method Plan security paths to avoid static and dynamic obstacles.
It enhances the realism and immersion of the virtual reality experience, ensuring that the drone avoids obstacles during flight, achieving safe flight and precise landing.
Smart Images

Figure CN119002359A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of virtual reality, and in particular to a VR immersive control system for controlling unmanned aerial vehicles. Background Art
[0002] Although the combination of VR (Virtual Reality) immersive technology and drones has been widely used in many fields and has shown great potential, they still face some technical challenges. One of the major shortcomings is the lack of an effective tactile feedback mechanism, which results in users being able to obtain a rich visual experience but not being able to obtain a corresponding supplement in terms of touch, thus weakening the overall immersive realism; in addition, in scenes where users are required to stand and participate, the inconsistency between virtual vision and the actual physical environment may also bring safety hazards; similarly, existing drone technology is prone to collisions and falls when autonomously navigating or being operated by non-professionals, which will not only damage the drone itself, but may also cause discomfort to users and even induce motion sickness. These shortcomings urgently need to be improved through technological innovation. Summary of the invention
[0003] In view of the above situation, the present invention provides a VR immersive control system for drone control. In view of the problem that the existing VR system lacks an effective tactile feedback mechanism, which weakens the realism of the overall immersive feeling, the present invention combines a VR helmet, a drone, an experience seat and a tactile feedback system to provide feedback on the state of the drone, uses an adaptive weighted multi-sensor data fusion method to process data information from various sensors, and provides accurate tactile feedback to users according to the state of the drone, thereby enhancing the realism and immersion of the virtual reality experience; in view of the fact that the existing drone technology is prone to collisions and falls when autonomously navigating or being controlled by non-professionals, the present invention builds an autonomous obstacle avoidance system into the drone. The system adopts an obstacle avoidance path prediction method, which can efficiently plan a safe path from the starting point to the target point, comprehensively considering the influence of static and dynamic obstacles, and ensuring that the drone can avoid obstacles throughout the flight process, achieving safe flight and precise landing.
[0004] The present invention provides a VR immersive control system for drone control, comprising a VR helmet, an experience seat, a drone, a control handle and a tactile feedback system;
[0005] The VR helmet is a high-resolution VR helmet with built-in sensors and a display screen, providing an immersive view from a first-person perspective, and tracking head movement to update the view direction;
[0006] The experience seat provides different degrees of vibration and tilt functions according to the flight status of the drone. The tilt angle of the experience seat simulates the posture change of the drone, so that the user's body sensation is consistent with the state of the drone.
[0007] The drone is a multi-rotor drone with stable flight capability, which is controlled by a control handle and has a built-in high-resolution camera and an autonomous obstacle avoidance system;
[0008] The control handle controls the drone, including a trigger button and a rocker, for accurately controlling the flight altitude, direction and speed of the drone;
[0009] The tactile feedback system receives multi-source sensor data from the VR helmet, experience seat, drone and control handle, uses an adaptive weighted multi-sensor data fusion method to perform fusion analysis on the multi-source sensor data, controls the VR helmet and experience seat to simulate tactile effects, and provides force feedback and vibration feedback;
[0010] The autonomous obstacle avoidance system uses GPS positioning technology to determine the actual position of the drone. At the same time, it adopts an obstacle avoidance path prediction method to monitor the status of the drone. When the drone loses control or is not controlled in time, it autonomously avoids obstacles and lands on flat ground.
[0011] Furthermore, the adaptive weighted multi-sensor data fusion method specifically comprises the following steps:
[0012] Step S1: Calculate the eccentricity distance. The multi-source sensor data consists of data from m sensors, each sensor has n observations. For each sensor's observation, calculate the eccentricity distance relative to the average value. The formula used is as follows: ;
[0013] In the formula, is the index, Indicates Observations, is the mean of the observed values, is the eccentric distance;
[0014] Step S2: Consistency preprocessing: perform consistency preprocessing on the observation values of each sensor according to the size of the eccentricity distance to remove The observation with the largest deviation;
[0015] Step S3: Calculate the weighting factor. For the remaining observations after consistency preprocessing, the weighting factor is calculated according to the eccentricity distance. The formula used is as follows: ;
[0016] In the formula, is the number of observations removed, is the total number of observations, Indicates The weighting factor for each observation;
[0017] Step S4: First-level fusion estimation: weighted fusion is performed on the observation values of each sensor to obtain the first-level fusion estimation value of the sensor. The formula used is as follows: ;
[0018] In the formula, Indicates First-level fusion estimate of sensors;
[0019] Step S5: First-level fusion estimation summary, repeating steps S1 to S4 to obtain the first-level fusion estimation value of each sensor;
[0020] Step S6: Calculate the optimal weighting factor. Calculate the optimal weighting factor based on the first-level fusion estimation values of all sensors. The formula used is as follows: ;
[0021] In the formula, Indicates The variance of the sensor observations, is the number of sensors, is the optimal weighting factor;
[0022] Step S7: Secondary fusion estimation: Calculate the secondary fusion estimation values of all sensors according to the calculated optimal weighting factor. The formula used is as follows: ;
[0023] In the formula, represents the secondary fusion estimate;
[0024] Step S8: Minimum total mean square error, calculate the minimum total mean square error corresponding to the secondary fusion estimation, the formula used is as follows: ;
[0025] In the formula, is the number of observations to be fused, is the minimum total mean square error.
[0026] Furthermore, the obstacle avoidance path prediction method specifically comprises the following steps:
[0027] Step P1: Expand the obstacle area. The drone captures obstacles within its sight through a built-in high-resolution camera and generates an expansion area in the form of a grid around the obstacle, with each grid as a node.
[0028] Step P2: Target confirmation, confirm the starting node and target node of the drone's obstacle avoidance;
[0029] Step P3: Global path planning, using the improved global path planning algorithm to traverse all paths to the target node, and select the static global optimal path through global path planning;
[0030] Step P4: Local path planning, using the improved DWA algorithm to obtain local information and avoid dynamic obstacles and unknown obstacles;
[0031] Step P5: Determine whether the UAV has reached the target node. If the UAV has reached the target node, end the path planning; otherwise, repeat the local path planning.
[0032] Furthermore, in step P3, the improved global path planning algorithm specifically includes the following steps:
[0033] Step P31: Obstacle rate calculation, define the obstacle rate , represents the distribution of obstacles, and the formula used is as follows: ;
[0034] In the formula, represents the number of obstacle grids, and Indicates the horizontal and vertical coordinates of the starting node. and Indicates the horizontal and vertical coordinates of the target node;
[0035] Step P32: Evaluation function, introduce the obstacle rate into the evaluation function, and adjust the weight to enhance the adaptability and global optimization capabilities. The formula used is as follows: ;
[0036] In the formula, represents the cost of moving from the starting node to the current location, represents the heuristic value, Representation Node The evaluation function of
[0037] Step P33: Optimize the selection of child nodes, divide the nodes into parent nodes and child nodes, and divide the child nodes into three categories according to the relationship between the child nodes and the parent nodes: upper and lower child nodes, left and right child nodes, and other child nodes; when the obstacle is located at a specific position, delete the left and right child nodes or the upper and lower child nodes of the obstacle;
[0038] Step P34: Bidirectional smoothness optimization, setting safety distance For any child node in the path, calculate the vertical distance between the child node and the center point of the obstacle. ,if , then the path is not feasible; otherwise, it is a safe path;
[0039] Furthermore, in step P4, the improved DWA algorithm specifically includes the following steps:
[0040] Step P41: Attitude adjustment: adjust the attitude of the drone to ensure that it is consistent with the direction of the current path node. The formula used is as follows: ;
[0041] In the formula, is the angular velocity, , is the minimum angular acceleration, is the maximum angular acceleration, Adjust the time for the posture, is the posture adjustment function;
[0042] Step P42: velocity sampling constraint, using the DWA algorithm to plan the path and obtain a set of feasible velocities, which are subject to the constraints of dynamics constraints, motor dynamics constraints and the instantaneous velocity before collision;
[0043] Step P43: Fuzzy controller, using the fuzzy controller to adjust the weights in the evaluation function in real time to enhance the adaptability of the algorithm;
[0044] Step P44: Improved evaluation function, adding additional weights to the evaluation function of the DWA algorithm , and add the azimuth of the child node to distinguish obstacles and avoid them. The formula used is as follows: ;
[0045] In the formula, , and is the weight coefficient, is the linear velocity, is the azimuth, is the current acceleration value, represents the shortest distance between the planned path and the static obstacle, represents the shortest distance between the planned path and the dynamic obstacle, is the evaluation function.
[0046] The beneficial effects achieved by the present invention using the above scheme are as follows:
[0047] (1) In view of the problem that the existing VR system lacks an effective tactile feedback mechanism, which weakens the realism of the overall immersion, this solution combines VR helmets, drones, experience seats and tactile feedback systems to provide feedback on the status of the drone, and uses an adaptive weighted multi-sensor data fusion method to process the data information of various sensors. This method fuses multi-sensor data by calculating eccentricity distance, consistency preprocessing, and weighted fusion operations to ensure the accuracy and reliability of the output results, enhance the robustness and precision of the system, and provide users with accurate tactile feedback based on the status of the drone, thereby enhancing the realism and immersion of the virtual reality experience;
[0048] (2) In view of the fact that existing drone technology is prone to collisions and falls when autonomously navigating or being operated by non-professionals, this solution has a built-in autonomous obstacle avoidance system in the drone. The system uses an obstacle avoidance path prediction method to efficiently plan a safe path from the starting point to the target point, taking into account the impact of static and dynamic obstacles to ensure that the drone can avoid obstacles throughout the flight process, achieving safe flight and precise landing. For static obstacles, the system uses an improved global path planning algorithm to generate an expansion area around the obstacle to ensure that the path maintains a sufficient safety distance from the obstacle. For dynamic obstacles, the improved DWA (Dynamic Window Approach) algorithm is used to avoid moving obstacles by adjusting the speed and direction in real time. In addition, an attitude adjustment function and a fuzzy controller are introduced to further optimize the path planning, ensuring that the drone can make the best obstacle avoidance decision even in a complex and changing environment, achieving safe and efficient autonomous flight. BRIEF DESCRIPTION OF THE DRAWINGS
[0049] Figure 1 A schematic diagram of a VR immersive control system for drone control provided by the present invention;
[0050] Figure 2 To experience the seat structure diagram;
[0051] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION
[0052] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all 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.
[0053] Example 1, see Figure 1 and Figure 2 , a VR immersive control system for drone control provided by the present invention includes a VR helmet, an experience seat, a drone, a control handle and a tactile feedback system;
[0054] The VR helmet is a high-resolution VR helmet with built-in sensors and a display screen, providing an immersive view from a first-person perspective, and tracking head movement to update the view direction;
[0055] The experience seat provides different degrees of vibration and tilt functions according to the flight status of the drone. The tilt angle of the experience seat simulates the posture change of the drone, so that the user's body sensation is consistent with the state of the drone.
[0056] The drone is a multi-rotor drone with stable flight capability, which is controlled by a control handle. The drone has a built-in high-resolution camera and an autonomous obstacle avoidance system. The high-resolution camera provides a front view for the VR helmet.
[0057] The control handle controls the drone, including a trigger button and a rocker, for accurately controlling the flight altitude, direction and speed of the drone;
[0058] The tactile feedback system receives multi-source sensor data from the VR helmet, experience seat, drone and control handle, uses an adaptive weighted multi-sensor data fusion method to perform fusion analysis on the multi-source sensor data, controls the VR helmet and experience seat to simulate tactile effects, and provides force feedback and vibration feedback;
[0059] The autonomous obstacle avoidance system uses GPS positioning technology to determine the actual position of the drone. At the same time, it adopts an obstacle avoidance path prediction method to monitor the status of the drone. When the drone loses control or is not controlled in time, it autonomously avoids obstacles and lands on flat ground.
[0060] Embodiment 2: This embodiment is based on the above embodiment. The experience seat provides different degrees of vibration and tilt functions according to the flight status of the drone, including the following simulation situations:
[0061] Acceleration / deceleration: When the drone accelerates, the experience seat can slightly push the user forward to simulate the feeling of being pushed back; when decelerating, the opposite happens, slightly pulling the user's body backward;
[0062] Ascending / descending: When ascending, the chair can be lifted slightly upward to simulate the feeling of lifting; when descending, it can be pressed down slightly to simulate the feeling of increased gravity;
[0063] Turning / Sliding: When turning left or right, the chair can tilt to the left or right accordingly to simulate lateral force; when sliding, the side motors of the chair can increase vibration to simulate lateral thrust;
[0064] Bumping / Turbulence: When encountering wind or air instability, the chair can produce random and slight vibrations to simulate the feeling of bumping; when simulating stronger turbulence, the vibration intensity and frequency can be significantly increased to reflect stronger bumps;
[0065] Stable flight: In a stable flight state, the experience seat can maintain slight vibration and provide continuous comfort;
[0066] Emergency: When the drone encounters a malfunction or emergency, the chair can sound an alarm and increase the overall vibration intensity to alert the user; at the same time, the chair can automatically enter protection mode to reduce unnecessary movements and ensure user safety.
[0067] Embodiment 3, based on the above embodiment, the adaptive weighted multi-sensor data fusion method specifically includes the following steps:
[0068] Step S1: Calculate the eccentricity distance. The multi-source sensor data consists of data from m sensors, each sensor has n observations. For each sensor's observation, calculate the eccentricity distance relative to the average value. The formula used is as follows: ;
[0069] In the formula, is the index, Indicates Observations, is the mean of the observed values, is the eccentric distance;
[0070] Step S2: Consistency preprocessing: perform consistency preprocessing on the observation values of each sensor according to the size of the eccentricity distance to remove The observation with the largest deviation;
[0071] Step S3: Calculate the weighting factor. For the remaining observations after consistency preprocessing, the weighting factor is calculated according to the eccentricity distance. The formula used is as follows: ;
[0072] In the formula, is the number of observations removed, is the total number of observations, Indicates The weighting factor for each observation;
[0073] Step S4: First-level fusion estimation: weighted fusion is performed on the observation values of each sensor to obtain the first-level fusion estimation value of the sensor. The formula used is as follows: ;
[0074] In the formula, Indicates First-level fusion estimate of sensors;
[0075] Step S5: First-level fusion estimation summary, repeating steps S1 to S4 to obtain the first-level fusion estimation value of each sensor;
[0076] Step S6: Calculate the optimal weighting factor. Calculate the optimal weighting factor based on the first-level fusion estimation values of all sensors. The formula used is as follows: ;
[0077] In the formula, Indicates The variance of the sensor observations, is the number of sensors, is the optimal weighting factor;
[0078] Step S7: Secondary fusion estimation: Calculate the secondary fusion estimation values of all sensors according to the calculated optimal weighting factor. The formula used is as follows: ;
[0079] In the formula, represents the secondary fusion estimate;
[0080] Step S8: Minimum total mean square error, calculate the minimum total mean square error corresponding to the secondary fusion estimation, the formula used is as follows: ;
[0081] In the formula, is the number of observations to be fused, is the minimum total mean square error.
[0082] In view of the problem that the existing VR system lacks an effective tactile feedback mechanism, which weakens the realism of the overall immersion, this solution combines VR helmets, drones, experience seats and tactile feedback systems to provide feedback on the status of the drone, and uses an adaptive weighted multi-sensor data fusion method to process data information from various sensors. This method fuses multi-sensor data by calculating eccentricity distance, consistency preprocessing, and weighted fusion operations to ensure the accuracy and reliability of the output results, enhance the robustness and precision of the system, and provide users with accurate tactile feedback according to the status of the drone, thereby enhancing the realism and immersion of the virtual reality experience.
[0083] Embodiment 4: This embodiment is based on the above embodiment, and the obstacle avoidance path prediction method specifically includes the following steps:
[0084] Step P1: Expand the obstacle area. The drone captures obstacles within its sight through a built-in high-resolution camera and generates an expansion area in the form of a grid around the obstacle, with each grid as a node.
[0085] Step P2: Target confirmation, confirm the starting node and target node of the drone's obstacle avoidance;
[0086] Step P3: Global path planning, using the improved global path planning algorithm to traverse all paths to the target node, and select the static global optimal path through global path planning;
[0087] Step P4: Local path planning, using the improved DWA algorithm to obtain local information and avoid dynamic obstacles and unknown obstacles;
[0088] Step P5: Determine whether the UAV has reached the target node. If the UAV has reached the target node, end the path planning; otherwise, repeat the local path planning.
[0089] Embodiment 5: This embodiment is based on the above embodiment. In step P3, the improved global path planning algorithm specifically includes the following steps:
[0090] Step P31: Obstacle rate calculation, define the obstacle rate , represents the distribution of obstacles, and the formula used is as follows: ;
[0091] In the formula, represents the number of obstacle grids, and Indicates the horizontal and vertical coordinates of the starting node. and Indicates the horizontal and vertical coordinates of the target node;
[0092] Step P32: Evaluation function, introduce the obstacle rate into the evaluation function, and adjust the weight to enhance the adaptability and global optimization capabilities. The formula used is as follows: ;
[0093] In the formula, represents the cost of moving from the starting node to the current location, represents the heuristic value, Representation Node The evaluation function of
[0094] Step P33: Optimize the selection of child nodes, divide the nodes into parent nodes and child nodes, and divide the child nodes into three categories according to the relationship between the child nodes and the parent nodes: upper and lower child nodes, left and right child nodes, and other child nodes; when the obstacle is located at a specific position, delete the left and right child nodes or the upper and lower child nodes of the obstacle;
[0095] Step P34: Bidirectional smoothness optimization, setting safety distance For any child node in the path, calculate the vertical distance between the child node and the center point of the obstacle. ,if , then the path is not feasible; otherwise, it is a safe path.
[0096] Embodiment 6: This embodiment is based on the above embodiment. In step P4, the improved DWA algorithm specifically includes the following steps:
[0097] Step P41: Attitude adjustment: adjust the attitude of the drone to ensure that it is consistent with the direction of the current path node. The formula used is as follows: ;
[0098] In the formula, is the angular velocity, , is the minimum angular acceleration, is the maximum angular acceleration, Adjust the time for the posture, is the posture adjustment function;
[0099] Step P42: velocity sampling constraint, using the DWA algorithm to plan the path and obtain a set of feasible velocities, which are subject to the constraints of dynamics constraints, motor dynamics constraints and the instantaneous velocity before collision;
[0100] Step P43: Fuzzy controller, using the fuzzy controller to adjust the weights in the evaluation function in real time to enhance the adaptability of the algorithm;
[0101] Step P44: Improved evaluation function, adding additional weights to the evaluation function of the DWA algorithm , and add the azimuth of the child node to distinguish obstacles and avoid them. The formula used is as follows: ;
[0102] In the formula, , and is the weight coefficient, is the linear velocity, is the azimuth, is the current acceleration value, represents the shortest distance between the planned path and the static obstacle, represents the shortest distance between the planned path and the dynamic obstacle, is the evaluation function.
[0103] Embodiment 7, this embodiment is based on the above embodiment, in step P42, dynamic constraint, the formula used is as follows: ;
[0104] In the formula, and are the minimum angular velocity and the maximum angular velocity respectively, is the dynamic constraint;
[0105] Motor dynamics constraints: ;
[0106] In the formula, and At any time The actual linear and angular velocities, and is the maximum acceleration and deceleration of the actual linear velocity, and is the maximum acceleration and deceleration of the actual angular velocity, represents the time increment, is the motor dynamics constraint;
[0107] Limitation of instantaneous speed before collision: ;
[0108] In the formula, Represents the current The shortest distance between the corresponding trajectory and the obstacle, is the limit of the instantaneous speed before the collision.
[0109] In view of the fact that existing UAV technology is prone to collision and falling when autonomously navigating or operated by non-professionals, this solution has a built-in autonomous obstacle avoidance system in the UAV. The system adopts an obstacle avoidance path prediction method and can efficiently plan a safe path from the starting point to the target point. It comprehensively considers the influence of static and dynamic obstacles to ensure that the UAV can avoid obstacles throughout the flight process and achieve safe flight and precise landing. For static obstacles, the system adopts an improved global path planning algorithm to generate an extended area around the obstacle to ensure that the path maintains a sufficient safety distance from the obstacle. For dynamic obstacles, the improved DWA algorithm is used to avoid moving obstacles by adjusting the speed and direction in real time. In addition, the attitude adjustment function and fuzzy controller are introduced to further optimize the path planning to ensure that the UAV can make the best obstacle avoidance decision even in a complex and changing environment, and achieve safe and efficient autonomous flight.
[0110] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article 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 process, method, article or device.
[0111] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.
Claims
1. A VR immersive control system for drone control, characterized in that: Including VR helmet, experience chair, drone, control handle and tactile feedback system; The VR helmet is a high-resolution VR helmet with built-in sensors and a display screen, providing an immersive view from a first-person perspective, and tracking head movement to update the view direction; The experience seat provides different degrees of vibration and tilt functions according to the flight status of the drone. The tilt angle of the experience seat simulates the posture change of the drone, so that the user's body sensation is consistent with the state of the drone. The drone is a multi-rotor drone with stable flight capability, which is controlled by a control handle and has a built-in high-resolution camera and an autonomous obstacle avoidance system; The control handle controls the drone to accurately control the flight altitude, direction and speed of the drone; The tactile feedback system receives multi-source sensor data from the VR helmet, experience seat, drone and control handle, uses an adaptive weighted multi-sensor data fusion method to perform fusion analysis on the multi-source sensor data, controls the VR helmet and experience seat to simulate tactile effects, and provides force feedback and vibration feedback; The autonomous obstacle avoidance system uses GPS positioning technology to determine the actual position of the drone. At the same time, it adopts an obstacle avoidance path prediction method to monitor the status of the drone. When the drone loses control or is not controlled in time, it autonomously avoids obstacles and lands on flat ground.
2. A VR immersive control system for drone control according to claim 1, characterized in that: The adaptive weighted multi-sensor data fusion method specifically comprises the following steps: Step S1: Calculate the eccentricity distance. The multi-source sensor data consists of data from m sensors, each sensor has n observations. For each sensor's observation, calculate the eccentricity distance relative to the average value. Step S2: Consistency preprocessing: perform consistency preprocessing on the observation values of each sensor according to the size of the eccentricity distance to remove The observation with the largest deviation; Step S3: Calculate the weighting factor. For the remaining observations after consistency preprocessing, the weighting factor is calculated according to the eccentricity distance. The formula used is as follows: ; In the formula, is the index, is the eccentric distance, is the number of observations removed, is the total number of observations, Indicates The weighting factor for each observation; Step S4: First-level fusion estimation, weighted fusion is performed on the observation value of each sensor to obtain the first-level fusion estimation value of the sensor; Step S5: First-level fusion estimation summary, repeating steps S1 to S4 to obtain the first-level fusion estimation value of each sensor; Step S6: Calculate the optimal weighting factor. Calculate the optimal weighting factor based on the first-level fusion estimation values of all sensors. The formula used is as follows: ; In the formula, Indicates The variance of the sensor observations, is the number of sensors, is the optimal weighting factor; Step S7: Secondary fusion estimation, calculating the secondary fusion estimation values of all sensors according to the calculated optimal weighting factors; Step S8: Minimum total mean square error, calculating the minimum total mean square error corresponding to the secondary fusion estimation.
3. A VR immersive control system and system for drone control according to claim 1, characterized in that: The obstacle avoidance path prediction method specifically comprises the following steps: Step P1: Expand the obstacle area. The drone captures obstacles within its sight through a built-in high-resolution camera and generates an expansion area in the form of a grid around the obstacle, with each grid as a node. Step P2: Target confirmation, confirm the starting node and target node of the drone's obstacle avoidance; Step P3: Global path planning, using the improved global path planning algorithm to traverse all paths to the target node, and select the static global optimal path through global path planning; Step P4: Local path planning, using the improved DWA algorithm to obtain local information and avoid dynamic obstacles and unknown obstacles; Step P5: Determine whether the UAV has reached the target node. If the UAV has reached the target node, end the path planning; otherwise, repeat the local path planning.
4. A VR immersive control system and system for drone control according to claim 3, characterized in that: In step P3, the improved global path planning algorithm specifically includes the following steps: Step P31: Obstacle rate calculation, define the obstacle rate , represents the distribution of obstacles, and the formula used is as follows: ; In the formula, represents the number of obstacle grids, and Indicates the horizontal and vertical coordinates of the starting node. and Indicates the horizontal and vertical coordinates of the target node; Step P32: Evaluation function, introduce the obstacle rate into the evaluation function and adjust the weight. The formula used is as follows: ; In the formula, represents the cost of moving from the starting node to the current location, represents the heuristic value, Representation Node The evaluation function of Step P33: Optimize the selection of child nodes, divide the nodes into parent nodes and child nodes, and divide the child nodes into three categories according to the relationship between the child nodes and the parent nodes: upper and lower child nodes, left and right child nodes, and other child nodes; when the obstacle is located at a specific position, delete the left and right child nodes or the upper and lower child nodes of the obstacle; Step P34: Bidirectional smoothness optimization, setting safety distance For any child node in the path, calculate the vertical distance between the child node and the center point of the obstacle. ,if , then the path is not feasible; otherwise, it is a safe path.
5. A VR immersive control system and system for drone control according to claim 3, characterized in that: In step P4, the improved DWA algorithm specifically includes the following steps: Step P41: Attitude adjustment: adjust the posture of the drone to ensure that it is consistent with the direction of the current path node; Step P42: velocity sampling constraint, using the DWA algorithm to plan the path and obtain a set of feasible velocities, which are subject to the constraints of dynamics constraints, motor dynamics constraints and the instantaneous velocity before collision; Step P43: fuzzy controller, using the fuzzy controller to adjust the weights in the evaluation function in real time; Step P44: Improved evaluation function, adding additional weights to the evaluation function of the DWA algorithm , and add the azimuth of the child node to distinguish obstacles and avoid them. The formula used is as follows: ; In the formula, , and is the weight coefficient, is the linear velocity, is the angular velocity, is the azimuth, is the current acceleration value, represents the shortest distance between the planned path and the static obstacle, represents the shortest distance between the planned path and the dynamic obstacle, is the evaluation function.
Citation Information
Cited By
Mixed reality flight training motion simulation system for virtually constructing scene
CN119942877A
