CN114779829B - A behavior control method and control system for a micro flapping-wing flying robot - Google Patents

Abstract

The invention belongs to the field of robot control, in particular to a behavior control method and a behavior control system for a miniature ornithopter flying robot, wherein a main end controller and a plurality of robot motion controllers are provided with intelligent control functions; and the master end controller sends the robot behavior instruction, and after the slave end robot motion controller receives the behavior level instruction, the master end controller analyzes and executes the behavior instruction to realize the robot behavior control. The invention saves the workload of operators, and the control track is stable and smooth. The control channel ratio of the behavior control mode is below 1%, so that the traffic and the control channel ratio are greatly reduced.

Description

A behavior control method and control system for a micro flapping-wing flying robot

Technical Field

The present invention belongs to the field of robot control, and in particular relates to a behavior control method and a control system for a micro flapping-wing flying robot.

Technical Background

Flapping-wing flying robots have broad application prospects in the field of UAVs due to their small size, light weight and good concealment. At the same time, flapping-wing flying robots can complete tasks such as environmental detection with high efficiency and over a large range. Micro flapping-wing flying robot clusters have great research value and significance in both military and civilian fields.

However, the traditional manual control method is to send the speed and angle control instructions of the flying robot in real time, and the control channel accounts for a high proportion. Since wireless communication is susceptible to interference, it is prone to problems such as poor signal, signal interruption, communication delay, signal blocking, etc. The flying robot cannot receive control instructions, which can easily lead to a series of accidents such as flying robots colliding, hitting obstacles, and aircraft losing control. This is also a problem that needs to be urgently solved for flapping-wing flying robots and rotorcraft robots.

Summary of the invention

The technical problem to be solved by the present invention is to provide a behavior control method and control system for a micro flapping-wing flying robot, so as to solve the problems of high channel occupancy and heavy operator workload in the prior art, and reduce the adverse effects on the control system caused by problems such as poor wireless communication, signal interruption, communication delay, and signal blocking.

The implementation process of the present invention is as follows:

A micro flapping-wing flying robot behavior control system, comprising:

The master-end controller and the slave-end robot motion controller, the robot motion controller has an intelligent control function, the master-end controller sends robot behavior instructions, and the slave-end robot motion controller receives the behavior-level instructions and parses and executes the behavior instructions to achieve robot behavior control. The master-end controller monitors the robot status in real time through robot vision and sensors, adjusts the robot status through behavior control instructions, and completes the task;

The master controller generates a motion behavior function sequence using a motion behavior generator according to the robot's task, and sends the motion behavior function control sequence to the robot motion controller to control the robot's actuator to complete the robot's task.

Furthermore, the robot motion controller includes a motion behavior function parser for parsing the motion behavior function control sequence; and includes a terminal motion control terminal for completing the task according to the parsed instructions.

Furthermore, the motion behavior function control sequence is composed of functions in a motion behavior function alphabet, and the motion behavior function alphabet is formulated according to actual motion characteristics of each motion behavior function.

Furthermore, the master-end controller generates a control logic based on a behavior tree according to the flapping-wing flying robot task, and generates a set of motion behavior functions; the behavior tree includes sequence nodes, fallback nodes, node conditions and robot behaviors, and the behavior tree includes control nodes and execution nodes. The control node is located inside the behavior tree and is used for logical reasoning and route navigation; the execution node is located at the bottom of the behavior tree and is used to judge conditions and execute actions; the control node includes sequence nodes and fallback nodes, and the execution node includes condition nodes and action nodes; during the execution of the behavior tree, the root node sends a frame signal to the lower-level nodes, transmits the signal downward in a left-to-right and top-to-bottom order, and executes each node in the tree. After the execution is completed, the execution result is fed back to the upper-level nodes layer by layer.

A behavior control method for a micro flapping-wing flying robot includes: performing path planning according to the robot's tasks, generating a motion behavior function sequence for a robot motion controller to control the robot's actuator to complete the robot's planning and operation tasks.

Furthermore, a motion behavior function control sequence is generated according to the flight robot task, and a control logic based on a behavior tree is generated. The control logic is determined by the tree structure of the behavior tree. The behavior tree includes a control node and an execution node. The control node is located inside the behavior tree and is used for logical reasoning and route navigation. The execution node is located at the bottom of the behavior tree and is used to judge conditions and execute actions. The control node includes a sequence node and a fallback node. The execution node includes a condition node and an action node. During the execution of the behavior tree, the control node sends a frame signal to the execution node at a certain frequency, transmits the signal downward in a depth-first order, and executes the execution nodes in the tree. After the execution is completed, the fallback node feeds back the execution result to the upper node layer by layer.

Furthermore, a behavior tree is generated according to the flapping-wing flying robot task, and the node conditions and action behaviors at the bottom of the behavior tree are combined into the control logic of the flapping-wing flying robot. The bottom-level node actions are combined into a behavior function set, and the behavior function and function set are used to control the slave robot to realize the behavior control of the flapping-wing flying robot.

Furthermore, in an emergency, the behaviors of the flapping-wing flying robot include direction adjustment, obstacle avoidance, base return, and communication restart. The behavior tree consists of sequence nodes and fallback nodes, as well as multiple node conditions and actions under the sequence nodes and fallback nodes. The node actions execute direction adjustment, obstacle avoidance, base return, or communication restart according to different node conditions of the flapping-wing flying robot.

Furthermore, the design method for obtaining each motion behavior function according to the actual motion characteristics of the flying robot behavior set is:

Use the manual control mode of the flapping-wing flying robot to complete the flapping-wing flying robot's operating tasks according to the actual operating task requirements, and record the gyroscope data and flight data in the manual control mode;

Fuzzy clustering analysis is performed on the gyroscope data to obtain the gyroscope data classification. The flight characteristics are summarized based on the data classification, and the motion behavior function of the corresponding category is designed.

Furthermore, the task decomposition process based on the behavior tree is as follows: according to the actual meaning of the total task, the total task is decomposed into the next layer of subtasks, and the next layer of subtasks is further decomposed layer by layer according to the type of task to form a tree structure, with the bottom layer being node conditions and node actions, and the backoff node is used to backoff the successfully executed tasks layer by layer to the upper layer, and all the child nodes of all the sequence nodes of the return layer are executed in sequence, and then backoff is used to feed back to the upper layer until the total task is completed; the sequence nodes of the same layer are executed in the order of the set tasks, wherein the execution principle of the backoff node is: execute each node from left to right and from top to bottom, and look for nodes that return successfully or are in the running state, until the total task is completed.

Compared with the prior art, the present invention has the following beneficial effects:

The present invention compares the manual control mode and the motion behavior control mode through experiments to observe their shortcomings and advantages. In the manual control mode controlled by the handle controller, the operator makes real-time judgments through the robot visual feedback, and the motion trajectory has the shortcomings of jitter and non-smoothness. The manual control mode requires the operator to make real-time judgments and controls, and the operator's workload is large and easy to fatigue. In the motion behavior control mode, the operator only needs to monitor and interact with a small number of robot instructions, which greatly saves the operator's workload and controls the trajectory smoothly. Therefore, the behavior-based flapping-wing flying robot control method has obvious advantages. As can be seen from Table 5, the control channel occupancy rate of the manual control mode is 100%, and control commands need to be sent in real time, while the control channel occupancy rate of the behavior control mode is 0.05%, which greatly reduces the communication volume. The operator's working time in the manual control mode includes monitoring time and control (issuing commands) time, and the workload ratio is 100%. The motion behavior control saves the time of issuing commands, so the operator's workload is 50%. Therefore, it can be seen from the simulation results that the control method based on motion behavior is not only effective but also feasible compared with manual control.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a block diagram of a system according to an embodiment of the present invention;

FIG2 is a behavior tree of a battlefield reconnaissance mission performed by a flapping-wing flying robot provided by an embodiment of the present invention;

FIG3 is a diagram of gyroscope data processed by the fuzzy C-means clustering method provided by an embodiment of the present invention;

FIG4 is a behavior tree diagram of a flapping-wing flying robot in an emergency obstacle avoidance situation provided by an embodiment of the present invention;

FIG5 is a spatial motion diagram of a flying robot based on manual control provided by an embodiment of the present invention;

FIG6 is a spatial motion diagram of a flying robot based on behavior control provided by an embodiment of the present invention.

DETAILED DESCRIPTION

In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

As shown in FIG1 , a micro flapping-wing flying robot behavior control system includes:

The master controller and the robot motion controller, the robot motion controller has intelligent control function. The master controller sends the robot behavior instructions, and after receiving the behavior level instructions, the robot motion controller of the slave end parses and executes the behavior instructions to realize the robot behavior control. The master controller monitors the robot status in real time through robot vision and sensors, adjusts the robot status through behavior control instructions, and completes the task.

The master controller performs path planning according to the robot's task, generates a trajectory according to the planned path, uses the motion behavior generator to generate a motion behavior function sequence, and sends the motion behavior function control sequence to the robot motion controller to control the robot's actuator to complete the robot's planning and operation tasks.

The robot motion device has some intelligent control functions, such as emergency obstacle avoidance. The master controller sends robot behavior instructions, and the slave receives the behavior-level instructions and parses and executes the behavior instructions to achieve robot behavior control. The master controller can control multiple slave robot motion controllers to achieve a distributed robot intelligent control mode. The master controller uses onboard vision and sensor data to make real-time judgments on the robot status, and adjusts the slave robot status through behavior control instructions.

The hardware system structure of the motion behavior control framework is shown in Figure 1. The functions in the motion behavior alphabet are in the form of L(v,ψ), allowing each function to be executed in an arbitrarily long period of time. In the system structure shown in Figure 1, the robot controller performs path planning and trajectory generation according to the task. Then, the motion behavior sequence is generated by the motion behavior function generator. Subsequently, the motion behavior control sequence is sent to the motion behavior parser to control the robot actuator to complete the robot's planning and operation tasks.

Robot behavior is formally defined as a symbolic string consisting of motion functions, and the symbolic string consisting of behavior is called motion planning.

A ground control unit is established with the master controller as the ground control station, and the behavior function parameters are sent to the flapping-wing robot to realize the behavior-level control command reception and command solution, thus forming the flapping-wing robot behavior movement.

The motion behavior control sequence is composed of functions in a motion behavior alphabet, which is defined according to a task-oriented robot behavior tree.

The main controller generates a control logic based on the behavior tree according to the flying robot task, and generates a motion behavior control sequence. The behavior tree is composed of sequence nodes, fallback nodes, node conditions and robot behaviors. The behavior tree includes control nodes and execution nodes. The control node is located inside the behavior tree and is used for logical reasoning and route navigation. The execution node is located at the bottom of the behavior tree and is used to judge conditions and execute actions. The control node includes sequence nodes and fallback nodes, and the execution node includes node conditions and execution actions. During the execution of the behavior tree, the root node sends a frame signal to each node, transmits the signal downward in order from left to right and from top to bottom, and executes the nodes in the tree. After the execution is completed, the execution results are fed back to the upper nodes layer by layer.

A behavior tree is a hierarchical, modular tree structure with a root node. It is usually used to express the behavior model of an intelligent agent. It is an effective method for describing the switching between different tasks in an autonomous intelligent agent or robot. Table 1 shows the node types and descriptions of the behavior tree, and Table 2 is a node description table:

Table 1 Node types and descriptions of behavior trees

Table 2 Node Description Table

The behavior tree includes two types of nodes: control nodes and execution nodes. The control node is located inside the behavior tree and is used for logical reasoning and route navigation. The execution node is located at the bottom of the behavior tree and is used to judge conditions and execute actions. The control nodes used in the present invention include fallback nodes and sequence nodes, and the execution nodes include node conditions and execution actions. During the execution of the behavior tree, the root node (sequence node at the top level) sends a frame signal and executes the node, and transmits the signal downward and executes each node in the tree in the order from left to right and from top to bottom. After the execution is completed, the execution results are fed back to the upper nodes layer by layer. The behavior tree of the battlefield reconnaissance and attack mission of the flapping-wing flying robot is shown in Figure 2. Therefore, the flapping-wing flying robot forms a control system with node conditions-robot behavior as the control logic.

The control logic is as follows: start execution from the 1st layer sequence node 0, execute the fallback node 1-1, perform patrol behavior until the target is found, and return to the 2nd layer sequence node 1-1 for success.

Continue to execute the 2-layer sequence node 1-2, continue to execute the 3-layer fallback node 2-2, perform the direction adjustment behavior, until the attack angle is met, and return to node 2-2 successfully.

Execute the fallback node 2-3. After the attack is confirmed, attack again until the attack is confirmed to stop. The fallback node 2-3 is successful. Return to the 1-2 node is successful.

Continue to execute the 1-3 sequence nodes, execute the fallback node 2-4, and if it is less than the safe distance, perform the avoidance behavior until it is greater than the safe distance, and the fallback 2-4 node is successful.

Then go back to node 2-5. If the fuselage is damaged, the robot will exit the battlefield and complete the flying robot mission.

Therefore, through the above definition and analysis of the behavior tree of the flapping-wing flying robot, the behavior tree of the flying robot is composed of sequence nodes, fallback nodes, node conditions and robot behaviors, and the control logic is determined by the node conditions and robot behaviors, which can be dynamically adjusted and optimized according to the tasks of the flying robot.

The design of the motion behavior function of the flapping-wing flying robot based on the behavior tree includes:

According to the task of the flapping-wing flying robot, the behavior set of the flapping-wing flying robot based on the behavior tree can be obtained. As shown in Figure 2, the task of the flapping-wing flying robot is battlefield reconnaissance and attack tasks, and the underlying node actions can all be defined as the behavior of the flapping-wing flying robot, such as patrolling, adjusting direction, attacking and exiting the battlefield.

The robot behavior function table with behavioral characteristics is defined as shown in Table 3:

Table 3 Behavior function table of flapping-wing flying robot

Behavior Classification Motion behavior function 1 Patrol Behavior L(v ₁ ,ψ ₁ ) 2 Adjust orientation behavior L(v ₂ ,ψ ₂ ) 3 Aggressive behavior L(v ₃ ,ψ3) 4 Exit the battlefield L(v ₄ ,ψ ₄ ) 5 Avoidance behavior L(v ₅ ,ψ ₅ )

The two-dimensional simplified model of the flying robot state space can be expressed as

Where (x, y) represents the coordinates of the center point of the flying robot, θ is the azimuth angle of the flying robot relative to the x-axis, v represents the linear velocity of the robot, and γ is the driving angle of the robot. The control quantity of the robot is the robot linear velocity v and the driving angle γ. Where x = (x, y, θ) ^T is the system state, u = (v, γ) ^T is the control input, and y = (x, y, θ) ^T is the system output.

The meaning of the L(v,ψ) function is explained as follows:

When the flying robot approaches the desired straight line in the forward direction, the azimuth angle θ of the flying robot will tend to the angle α(·) of the desired straight line. The corresponding symbols in the function are defined as:

v ₁ =v′

φ ₁ =K tan ^-1 (k ₁ ·(k ₂ ·Δα(·)+k ₃ vω(Δα(·))δ(x,y)))

Where v ₀ is a constant, k ₁ , k ₂ and k ₃ are control gain constants, K is the control angle gain variable, Δα(·) is the difference between the desired angle and the actual angle, w(Δα) is the angle difference gain function, and δ(x, y) is the system state desired angle function. The symbolic definition and function description of L(v ₂ ,ψ ₂ ) and L(v ₃ ,ψ ₃ ) refer to L(v ₁ ,ψ ₁ ).

According to the definition of the above function, the motion alphabet of the robot motion, such as: ∑ = {L ₁ , L ₂ , L ₃ , L ₂ , L ₁ ......}, is used to optimize the behavior of the flapping-wing flying robot:

The defined flapping-wing flying robot behavior function table is optimized. First, each behavior function is analyzed, and the behavior functions that are difficult to execute and complete for the flapping-wing flying robot should be removed. Two or more behaviors can also be combined to construct a composite behavior that is conducive to task execution.

The design method for obtaining each motion behavior function based on the actual motion characteristics of the flapping-wing flying robot behavior set is:

Utilize the manual control mode of the flapping-wing flying robot to complete the flapping-wing flying robot's operating tasks according to the requirements of the actual operating tasks, and record the onboard gyroscope data and flight data in the manual control mode.

Fuzzy clustering analysis is performed on the gyroscope data to obtain the classification data of the gyroscope. The flight characteristics are extracted based on the classification data, and the motion behavior function of the corresponding category is designed.

Using on-site manual control experimental data, a characteristic behavior design method is designed for actual work tasks.

First, the flapping-wing flying robot manual control mode is used to complete the flapping-wing flying robot's operating tasks according to the requirements of the actual operating tasks, and the gyroscope data and flight data in the manual control mode are recorded. Fuzzy clustering analysis is performed on the gyroscope data to obtain the classification data of the gyroscope. Based on this, the flapping-wing flying robot behavior function and function set for the actual operating tasks are set, so that the flapping-wing flying robot's behavior function has more flight characteristics of the operating tasks, so as to realize a new method for designing the flying robot behavior function for the operating tasks, as shown in Figure 3.

Among them, the fuzzy clustering process is to perform fuzzy clustering on the three-dimensional data of the gyroscope, randomly select three initial points as the three cluster initial centers, calculate the distance from each point to the three cluster centers, and then process the distance with the fuzzy function to obtain the fuzzy value, calculate the fuzzy matrix from all points to the three cluster centers respectively, divide them into the nearest class according to the fuzzy value, and then calculate the new cluster center for the next cycle.

The construction process of the behavior tree is as follows: according to the actual meaning of the total task, the total task is decomposed into the next layer of subtasks, and the next layer of subtasks is further decomposed layer by layer according to the type of task to form a tree structure, with the bottom layer being node conditions and execution actions, to achieve task decomposition based on the behavior tree. The execution principle of the fallback node is: start executing each node from left to right and from top to bottom, and look for nodes that return success or are in the running state until the total task is completed.

According to the actual meaning of the task, the initial task is decomposed by using the success mode of a single child node of the fallback node and the success mode of all child nodes of the sequence node. Then, according to the actual meaning of the hierarchical subtasks, the nodes are decomposed into fallback nodes, sequence nodes, node conditions, and action behaviors, so as to meet the requirements of the task in terms of the actual meaning, and to use the node conditions and action behaviors to meet the actual operation process of the entire task cycle.

The present invention also provides a method for controlling the autonomous obstacle avoidance and fault tolerance behaviors of the robot in an emergency state.

When the flapping-wing flying robot is in an abnormal state such as poor communication signal, signal interruption, communication delay, signal blocking, etc., the fault operation mode will be started. Based on the description of the fault operation behavior tree, with safe obstacle avoidance and communication connection as subtasks, an autonomous obstacle avoidance operation mode of the flapping-wing flying robot in an emergency state is formed. The behavior function table of the flapping-wing flying robot is shown in Table 4, and the behavior tree is shown in Figure 4.

Table 4 Behavior function table of flapping-wing flying robot

Behavior Classification Motion behavior function 1 Adjust orientation behavior L(v ₁ ,ψ ₁ ) 2 Obstacle avoidance behavior L(v ₂ ,ψ ₂ ) 3 Return to base L(v ₃ ,ψ ₃ ) 4 Communication restart L(v ₄ ,ψ ₄ )

Through the above-mentioned autonomous obstacle avoidance and fault-tolerant control mode of the flapping-wing flying robot in emergency situations, the obstacle avoidance control of the flapping-wing flying robot in emergency situations can be realized through direction adjustment behavior, obstacle avoidance, base return and communication restart behavior, providing reliable robot behavior control support for the flapping-wing flying robot to return to the ground control station and reconnect communication.

The present invention provides a behavior control method for a micro flapping-wing flying robot, which constructs a behavior tree according to the robot's task, and generates a motion behavior sequence according to the behavior tree to give to the robot motion controller to complete the robot's operation task.

According to the task of the flapping-wing flying robot, a control logic based on the behavior tree is generated to generate a motion behavior control sequence. The behavior tree includes a control node and an execution node. The control node is located inside the behavior tree and is used for logical reasoning and route navigation. The execution node is located at the bottom of the behavior tree and is used to judge conditions and execute actions. The control node includes a sequence node and a fallback node. The execution node includes node conditions and execution actions. During the execution of the behavior tree, the root node sends a frame signal to each node, transmits the signal downward in the order from left to right and from top to bottom, and executes each node in the tree. After the execution is completed, each node will feedback the execution result to the upper node layer by layer.

The node action is the behavior of the flapping-wing flying robot. According to the task of the flapping-wing flying robot, the flying robot behavior set based on the behavior tree is obtained, and the motion behavior function control sequence is composed of functions in the motion behavior function alphabet.

In an emergency, the flapping-wing flying robot behavior function includes direction adjustment behavior, obstacle avoidance behavior, base return and communication restart behavior. The behavior tree uses sequence nodes and fallback nodes, as well as multiple node conditions and execution actions under the sequence nodes and fallback nodes. The node actions execute direction adjustment behavior, obstacle avoidance behavior, base return or communication restart behavior according to the flapping-wing flying robot behavior function.

The design method for obtaining the motion behavior function based on the actual motion characteristics of the flapping-wing flying robot is:

Utilize the manual control mode of the flapping-wing flying robot to complete the flapping-wing flying robot's operating tasks according to the requirements of the actual operating tasks, and record the gyroscope data and flight data in the manual control mode.

Fuzzy clustering analysis is performed on the gyroscope data to obtain the gyroscope data classification. The flight characteristics are summarized based on the data classification, and the motion behavior function of the corresponding category is designed.

The construction process of the behavior tree is as follows: according to the actual meaning of the total task, the total task is decomposed into the next layer of subtasks, and the next layer of subtasks is further decomposed layer by layer according to the type of task to form a tree structure, with the bottom layer being node conditions and execution actions, to achieve task decomposition based on the behavior tree. The execution principle of the fallback node is: start executing each node from left to right and from top to bottom, and look for nodes that return success or are in the running state until the total task is completed.

Verify the effectiveness of the solution of the embodiment of the present invention:

By comparing the manual control method and the motion behavior control method in the experiment, we can observe their disadvantages and advantages. In the manual control method controlled by the handle controller, the operator makes real-time judgments through the robot's visual feedback, and the robot trajectory has the disadvantages of jitter and non-smoothness, as shown in Figure 5. The manual control method requires the operator to perform real-time control and judgment, which is a heavy workload for the operator and easy to fatigue. In the motion behavior control method, the operator only needs to monitor and interact with a small number of robot commands, which greatly saves the operator's workload, and the control trajectory is smooth, as shown in Figure 6. Therefore, the behavior-based flapping-wing flying robot control method has obvious advantages.

Table 5 Comparison of control parameters between manual control mode and motion behavior mode

As can be seen from Table 5, the communication control channel occupancy rate of the manual control mode is 100%, and control commands need to be sent in real time, while the control channel occupancy rate of the behavior control mode is 0.05%, which greatly reduces the communication volume. The operator workload of the manual control mode includes monitoring time and control (issuing commands) time, and the workload ratio is 100%. The motion behavior control saves the time of issuing commands, and the operator workload is close to 50%. Therefore, from the simulation results, it can be seen that the control method based on motion behavior is not only effective but also feasible compared with manual control.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims (6)

1. A miniature ornithopter robot motion control system, comprising:

The robot comprises a master end controller and a slave end robot motion controller, wherein the robot motion controller has an intelligent control function, the master end controller sends a robot behavior instruction, the slave end robot motion controller analyzes and executes the behavior instruction after receiving the behavior level instruction to realize behavior control of a robot, the master end controller monitors the state of the robot in real time through a robot vision and a sensor, and the robot state is adjusted through the behavior control instruction to complete a task;

the master end controller generates a motion behavior function sequence by adopting a motion behavior generator according to the task of the robot, and sends the motion behavior function control sequence to the robot motion controller so as to control the robot executing mechanism to complete the operation task of the robot; the robot motion controller comprises a motion behavior function analyzer which analyzes a motion behavior function control sequence; the terminal motion control end is used for completing tasks according to the analyzed instructions;

The motion behavior function control sequence is composed of functions in a motion behavior function alphabet, and the motion behavior function alphabet is formulated according to actual motion characteristics of each motion behavior function;

the main end controller generates control logic based on a behavior tree according to the flapping wing flying robot task, and generates a motion behavior function set; the behavior tree comprises a sequence node, a rollback node, node conditions and robot behaviors, the behavior tree comprises a control node and an execution node, and the control node is positioned in the behavior tree and used for logic reasoning and route navigation; the execution node is positioned at the bottom end of the behavior tree and is used for judging conditions and executing actions; the control node comprises a sequence node and a rollback node, and the execution node comprises a condition node and an action node; in the execution process of the behavior tree, a root node sends out a frame signal to a lower node, signals are transmitted downwards according to the sequence from left to right and from top to bottom, each node in the tree is executed, and after the execution is finished, an execution result is fed back to an upper node layer by layer;

In an emergency state, the behavior of the ornithopter robot comprises a direction adjustment behavior, an obstacle avoidance behavior, a base return and a communication restarting behavior, wherein a behavior tree consists of a sequence node and a rollback node, and a plurality of node conditions and actions under the sequence node and the rollback node; the node action executes direction adjustment action, obstacle avoidance action, base return or communication restarting action according to different node conditions of the ornithopter.

2. A miniature ornithopter robot motion control method employing a miniature ornithopter robot motion control system as set forth in claim 1, comprising: and planning a path according to the task of the robot, and generating a motion behavior function sequence to a motion controller of the robot so as to control an executing mechanism of the robot to complete the planning and the operation task of the robot.

3. The control method according to claim 2, characterized in that a control sequence of a movement behavior function is generated according to the flying robot task, a control logic based on a behavior tree is generated, the control logic is determined by a tree structure of the behavior tree, the behavior tree comprises control nodes and execution nodes, and the control nodes are located inside the behavior tree and used for logic reasoning and route navigation; the execution node is positioned at the bottom end of the behavior tree and used for judging conditions and executing actions, and the control node comprises a sequence node and a rollback node; the execution node comprises a condition node and an action node, in the execution process of the action tree, the control node sends frame signals to the execution node according to a certain frequency, the signals are transmitted downwards according to the depth priority sequence, the execution node in the execution tree, and after the execution is finished, the rollback node feeds back the execution result to the upper node layer by layer.

4. A control method according to claim 3, characterized in that the behavior tree is generated according to the flapping-wing flying robot task, the control logic of the flapping-wing flying robot is formed by combining the node conditions and the action behaviors of the bottom layer of the behavior tree, the behavior function set is formed by combining the node actions of the bottom layer, and the behavior control of the flapping-wing flying robot is realized by controlling the slave robot by using the behavior function and the function set.

5. A control method according to claim 3, characterized in that the design method for obtaining each movement behavior function from the actual movement characteristics of the flying robot behavior set is:

The method comprises the steps of completing an operation task of the flapping-wing flying robot according to actual operation task requirements by using a manual control mode of the flapping-wing flying robot, and recording gyroscope data and flight data in the manual control mode;

And carrying out fuzzy cluster analysis on the gyroscope data to obtain data classification of the gyroscope, and designing a motion behavior function of a corresponding class according to the data classification and the flight characteristics.

6. A control method according to claim 3, characterized in that the task decomposition process based on the behavior tree is: decomposing the total task into a next layer of subtasks according to the actual meaning of the total task, decomposing the next layer of subtasks layer by layer according to the type of the task to form a tree structure, wherein the bottom layer is a node condition and a node action, feeding back the successfully executed task layer by layer to the upper layer by using a rollback node, sequentially executing all the subtasks of all sequence nodes of the returned layer, and feeding back the rest to the upper layer by rollback until the total task is completed; the sequence nodes of the same layer are executed according to the sequence of the set tasks, wherein the execution principle of the rollback nodes is as follows: and starting to execute each node from left to right and from top to bottom, searching for success or running state until the total task is completed.