Quadruped Tracked Robot

Preface: This project attempts to replicate the work of the tracked quadrupedal composite robot [Graduation Project], and conducts basic tests and improvements on the use of this machine in a series of terrains.
 
1. Project Introduction
  In this project, the quadrupedal tracked robot has three travel modes: quadrupedal travel mode, tracked travel mode, and tracked-quadrupedal collaborative mode. Thanks to the flexible mechanical arm structure, the body posture can be adjusted according to different terrains to optimize off-road performance. The foot end integrates the track structure, and each track structure is driven by a DC motor.
  The robot uses ESP32-WROVER-E as the main control core, which can realize the control of the bus servo and DC motor. The image transmission solution uses Raspberry Pi to drive the FHD camera module 4B to push the stream.
 The tracked vehicle form can maintain a higher travel speed on relatively flat terrain and has lower power consumption; the quadruped form is suitable for crossing obstacles in complex environments and has better performance; the tracked-quadrupedal collaborative mode has a higher travel speed and obstacle crossing ability, but has higher power consumption.
2. Electronic control hardware function
6-way serial bus interface: PH2.0-3P, 74HC125N
4-way DC motor drive: TB6612FNG
0.96-inch 4-pin OLED screen interface: HDR-1*4P-F
3. Selection description
Serial bus servo selection: Model: RA8-U25, rated voltage: 6.0~8.4V, serial communication baud rate 9600~1Mbps, rated torque 25kg/cm, effective angle 270°. Communication needs to convert UART full-duplex to half-duplex. According to the UART serial bus servo manual provided by FashionStar merchants, 74HC125N is used to achieve it; the initial test occasionally releases the servo in a complex environment, which is improved after replacing the servo. Model: RP8-U45, rated voltage: 6.0~8.4V, serial communication baud rate 9600~1Mbps, rated torque 45kg/cm, effective angle 360°. This version of the servo has a large power, which can fully improve the difficulty of traveling in complex terrain caused by low multi-level power. 3. DC motor selection: Model: JGA25-370-1285, rated voltage 12.0V, reduction ratio: 1:103, no-load speed: 62 rpm. In actual measurements, due to high load and insufficient power in complex terrain such as mountain gullies, it is possible to consider replacing a larger power motor.
4 Power supply selection: Use 3S polymer aircraft model battery with a discharge rate of 60C. If a low-power module is used, the discharge rate can be appropriately reduced.
4. Circuit design
1. Servo drive circuit: Use 74HC125N chip, use 8 serial bus servos, 4 interfaces are sufficient, and two more interfaces are added to prevent subsequent damage. The following is the servo drive circuit diagram.
2. Motor drive circuit: The drive chip is TB6612FNG. One chip can drive two motors. The peripheral circuit is simple and takes up less space.
3. Motion sensor: MPU6050 is used. QFN package is difficult to weld, so the module is used directly.
4. Power supply: 5V and 3.3V are realized through LDO, namely AMS1117-5V and AMS1117-3.3V. LD1117 series can also be used.
Since the bus servo needs 8V power supply, an 8V power supply is added. XL4016E1 adjustable power supply is used, 1.25-36V adjustable, if 8V is not needed, it can be deleted.
 
 
5. Actual picture Main
control board:  
 
Structure: The side connector of the fuselage has been thickened and reinforced. The model picture is as follows.
In order to strengthen the connection between the fuselage and the overall robotic arm, the leg drive connector has been lengthened. The second picture is the actual picture after the change. In order to avoid collision between the fuselage and the robotic arm servo during movement, the robotic arm connection part is changed to a stainless steel connector with a thickness of about 6mm. The track part uses a leopard metal track.
6. Basic Implementation Ideas
For the posture part, this attempt mainly verifies the implementation of Walk and Trot gaits under this model; the figure shows the motion status of each robot arm under Trot gait. It has been verified that this can be used as a basis to realize the basic Trot posture in this machine. The first half cycle of Tort (as shown in Figure 6-3): the left front leg and the right hind leg are in the swing phase, and the right front leg and the left hind leg are in the support phase. In the swing phase, the left front leg moves forward and the right hind leg moves backward; the support phase remains unchanged. This completes the first half cycle of Tort.
The second half cycle of Tort (as shown in Figure 6-4): the left front leg and the right hind leg are in the support phase, and the right front leg and the left hind leg are in the swing phase. In the swing, as shown in the figure below, the left front leg and the right hind leg are in the support phase, and the right front leg and the left hind leg are in the swing phase. In the swing term, the right front leg moves forward and the left hind leg moves forward. In the support term, the left front leg exerts force and pushes backward; the right hind leg exerts force and pushes backward, driving the vehicle body to move forward. Continue for the second half of the cycle; the second half of the Tort cycle is now completed.
For the obstacle avoidance part, (modification test in progress) basic detection of environmental obstacles is performed through sensors such as LiDAR, and basic target following is attempted with UWB. The body adaptation test is currently being modified...
7. The upper position
uses Raspberry Pi + Quectel EC20 module to complete streaming and communication, and is compatible with Bluetooth serial port communication and network TCP protocol.
6. Scene test
To verify the practicality of the body's functions, the stability and speed are tested in a variety of scenes such as stairs, gullies, rugged earth mounds, and flat roads: Climbing
stairs Climbing concrete roads Current progress: The body load is large, and some parts have a large negative pressure; some parts are being improved in the future to enhance strength and reduce the weight of the body. Redesign the foot structure to reduce the power transmission loss in the machine. Try to achieve balance control by linking the accelerometer and gyroscope to the robotic arm. Try to redesign the camera gimbal to optimize image transmission and basic image recognition.