Emergency Response via Spherical - Ducted UAV Integration

• Main Contributions:

  Core member of spherical robot software (non-flight part)

  1. Building a control system framework

graph TD
    subgraph Communication_System
        A[Remote_Controller] --> B[Wireless_Module]
    end
    subgraph Sensor_System
        C[Inertial_Measurement_Unit]
        D[Joint_Motor_Encoder]
    end
    subgraph Sphere_Execution_System
        E1[Joint_Motor_Driver_1] --> F1[Rolling_Motor]
        E2[Joint_Motor_Driver_2] --> F2[Pendulum_Motor]
    end
    subgraph Flight_System
        G[Gyroscope] --> H[Flight_Controller]
        H --> I[Brushless_Motor_Driver*4]
        I --> J[Duct_Motor*4]
    end
    B --> K[Main_Control_Chip]
    C --> K
    D --> K
    K --> E1
    K --> E2
    K --> H

                                        Figure 2. Control system flow chart

The system is composed of four parts, with the following functions:

1. Communication System: It consists of a remote controller and a wireless module. This system transmits relevant instructions to the main control chip, enabling the remote - control function of the robot.
2. Sensor System: Comprising an IMU and a joint motor encoder. It measures the object’s acceleration, angular velocity and other velocity - related data, and detects the position and speed of the joint motor. This provides real - time status information of the robot to the main control chip.
3. Sphere Execution System: Has two joint motors. These motors are used to drive the robot for rolling and swinging respectively.
4. Flight Control System: The Pixhawk 6C Mini is directly used for configuration. Through the flight control, it controls four ducted motors to achieve the flight control of the robot.

        2.Implement a rolling control algorithm

graph TD
    subgraph MPC_Controller
        A[Prediction Model] --> B[Objective Function]
        A --> C[Quadratic Programming]
        B --> C
        D[Constraints] --> B
        D --> C
        C --> E[Future Control Sequence]
        E --> A
    end
    F[Reference Trajectory] --> A
    G[Spherical Robot Dynamics] --> A
    G --> H[State Variables]
    H --> C
    E --> I[First Control Quantity]
    I --> G

                     Figure 3.Model prediction algorithm flow chart

1.Algorithm Architecture Selection: Chose the MPC algorithm architecture for the design.

2.Model Processing: Referred to relevant papers to simplify the dynamics and kinematics model of the sphere.

3.Prediction and Index Establishment: Utilized the method from the papers to combine the reference trajectory with the actual situation, established the performance index, and converted it into a quadratic optimization problem.

4.Algorithm Optimization: Employed rolling programming to evaluate the effect within the limited time domain, and continuously adjusted the control step size, prediction step size, and weight matrix to optimize the algorithm.