Robot | Benjamin

Trunk Robot

Wed, 04 May 2022 00:00:00 +0000

I am the general manager of the self-developed project - the bionic elephant trunk robot based on the wire pulling technology. I was responsible for personnel scheduling, task allocation and overall positioning design of the project. I was also responsible for embedded software development and hardware topology design.

👉 Github: https://github.com/BenjaminChen1024/Trunk-Robot

Trunk Robot

Our team designed the bionic mechanical elephant trunk ecological restoration robot, based on the bionic wire pulling technology and the bionic mechanical transmission principle of elephant trunk.

Trunk Robot

Mechanical Structure

In terms of mechanical design, the robot adopts modular design, including: land-air integrated chassis, 360-degree rotating head, multi-degree-of-freedom line driven elephant trunk and modular end actuator, so as to realize ground movement, climbing, rotation, flexible movement of elephant trunk, clamping, cutting and other functions.

Trunk Robot Mechanical Structure

Hardware

The power supply part is powered by 24V battery, and the chassis is driven by four 6020 motors to realize the free movement of front and back. The elephant trunk is controlled by a 6020 motor to control the YAW shaft of the cradle head and four 3508 motors to drive the wire pulling mechanism. The power supply independently supplies power to the STM32 main control board, which supplies power and controls the end-actuator steering gear. The power supply is supplied to the Raspberry PI through the 24V to 5V step down module, and the Raspberry PI supplies power to the camera through USB.

Trunk Robot Hardware

Softwork

The embedded underlying driver is developed based on STM32 chip, and FreeRTOS real-time operating system is used for task management. System tasks include: remote control communication, mode switching, dropped line protection, PC communication, motor control, steering gear control, debugging mode, system reset. Initialize the Flash, Timer, CAN, and UART.

Trunk Robot Softwork

Machine Vision

Visual module uses raspberry sent 4 b as operation platform, driven by industrial camera to capture frames flow using OpenCV vision library for feature extraction, and the data sent to the embedded microcontroller, visual AIDS control functions, through object capture and fixed position, realize automatic clearing up the rubbish, automatic aerial work, etc.

Trunk Robot Machine Vision

Product Display

Awards

Bionic Fish

Wed, 13 Apr 2022 00:00:00 +0000

I was in charge of the Bionic Fish project, responsible for the development of embedded systems and machine vision.

Bionic Fish

Although the traditional propulsion mode of “propeller + steering rudder” can better provide large thrust and high speed, its application in the above occasions is limited due to its inherent characteristics, such as obvious fluid disturbance and poor low-speed maneuvering performance. Bionic Fish can meet the characteristics of small turbulence, high propulsion efficiency, small wake characteristics and good maneuverability.

Bionic Fish Design

Model

The Bionic Fish model includes mechanical design and hardware deployment. Through the double structure with waterproof paint, can effectively improve the water - proofing of fish. The Bionic Fish moves freely through the water by swinging its fins and caudal fins.

Bionic Fish Model

Hardware

Hardware includes power supply, switch, sensor, controller, camera, drive element, indicator element. The stability of the hardware in water can be improved by designing the PCB and arranging the space properly.

Bionic Fish Hardware

Control

Bionic Fish through infrared sensor, machine vision parameters, MPU6020 Angle sensor as input, through the low-pass filter and PID controller output error parameters to control the steering gear rotation Angle, so as to control the posture and movement of the Bionic Fish.

Bionic Fish Contorl

Future

Because the steering gear of the original model is directly connected to drive the fin, there are certain shortcomings, such as large mass, large occupancy volume and slow response speed. The new Bionic Fish will be based on the line pull technology for the fish tail adaptive control, using a servo motor with faster response speed to drive.

New Bionic Fish Design

Robomaster - Engineer Robot

Sat, 04 Sep 2021 00:00:00 +0000

I am a member of engineering embedded software in the 21st season of Scut Robot Lab, responsible for engineering robot hardware deployment and embedded software research and development.

👉 Github: https://github.com/BenjaminChen1024/21_Engineer_Top

Engineer Robot

Engineer robot is responsible for competing for its own resources and resurrecting its own robot, which consists of two parts: mobile chassis and upper mechanism. The mobile chassis includes the McNamham wheel, the rescue mechanism and the card swiping mechanism which can move in all directions. The upper mechanism includes lifting mechanism, extending mechanism, flipping mechanism and clamping mechanism.

Engineer Robot Model

Hardware

Engineer robot hardware uses STM32F405 chip as the controller. Robomaster M3508 motor and 2006 motor are used as drivers. CAN communication protocol and serial communication protocol are adopted. Photoelectric gate and visual camera are used as sensors.

Engineer Robot Hardware

PCB Development

Engineer robot PCB has Chassis and Topper, which are responsible for controlling Chassis and Topper hardware devices respectively.

Chassis PCB consists of STM32F405 chip, IMU chip, CAN communication module, serial port communication module, IIC communication module, and DBUS communication module.

Engineer Robot Chassis PCB

Topper PCB is composed of STM32F405 chip, CAN communication module, serial communication module, IIC communication module, free IO port, overcurrent protection module and relay module.

Engineer Robot Topper PCB

Softwork

Engineer robot software is based on FreeRTOS real-time operating system, and the different modules of the robot are controlled by multi-thread management. These include motor control, visual assistance, mode switching, drop protection, handle key control and so on.

Engineer Robot Software

Robot Control

Engineer robot control is divided into action group and automatic group. The action group is assigned by the state machine to execute individual actions, including rotation, flip, extension, lift, cylinder control. The automatic group is to control the state machine through continuous instructions to execute corresponding continuous actions, including one-key mining, one-key exchange, one-key reset, automatic rescue, one-key interrupt protection.

Engineer Robot Control

Debug

In order to make the movement of the robot more stable and prevent various accidents in the competition, we need to debug the robot again and again, and record its test parameters such as: clamping success rate, exchange success rate, chassis moving speed, etc.

Competition

Through continuous iteration and debugging, we finally won the victory and won the first prize of the National College Student Robotics Competition RoboMaster2021 Mecha Master Competition.

My engineering robot and I