K230 Target Tracking

This tutorial is a comprehensive experiment combining multiple peripherals. It is recommended to understand individual peripherals before proceeding with this experiment.

1. Software-Hardware Requirements

• KEIL5

• MSPM0G3507 Robot Development Board

  K230 module: External connection

• Type-C Data Cable or DAP-Link

  For program download or simulation to the development board

2. Basic Principles

2.1 Hardware Schematic Diagram

2.2 Physical Connection Diagram

Wiring Pins

Motor Wiring (Note: The wiring in the diagram below is for position reference only. Our factory provides dual-head PH2.0 6Pin all-black cables with anti-mistake design, so you don't need to worry about wiring issues)

K230 Wiring (Note: The wiring in the diagram below is for position reference only. Our factory provides K230 dual-head PH2.0 4Pin all-black cables with anti-mistake design, so you don't need to worry about wiring issues)

2.3 Control Principle

• Program Flowchart

Control Principle:

PID Control

PID control is a commonly used closed-loop control algorithm that continuously adjusts the controller output to minimize the error between the system's actual output and expected output.

Communication Protocol:

When K230 recognizes the learned object, it will send a data frame via serial port with the following format:

Microcontroller part:

Receives data sent by K230 via serial port and processes each byte of data. Extracts valid data from the data frame content and stores it in an array. Then performs PID processing on the array data, and the microcontroller drives the motor movement using the processed data to move along the track.

3. Code Analysis

Main Function Analysis

Function: Pto_Loop

Partial Function Analysis

• main.c

​x
​
ColorMode color_mode = 2;  //1: Patrol line, 2: Tracking
​
​
int main(void)
{
    
    bsp_init(); // Required peripheral initialization
    Control_RGB_ALL(OFF); // Turn off RGB
    delay_ms(100);
    OLED_ShowString(0,0,"pid_output:",8,1); // Display PID output for left-right control
    OLED_ShowString(0,20,"pid_output1:",8,1); // Display PID output for forward-backward control
    OLED_Refresh();
            
    while (1) 
    {                   
            Pto_Loop();
            OLED_ShowSNum(75,0, pid_output, 3, 8, 1);
            OLED_ShowSNum(75,20,pid_output1,3,8,1);
            OLED_Refresh();
            
    }
}
​

color_mode: Use 1 when running vision patrol line case, use 2 when running tracking case

BSP_init: Hardware initialization

pto__Loop: Continuously processes received K230 messages.

• yb_protocol.c

xxxxxxxxxx
/**
 * @Brief: 数据分析 Data analysis
 * @Note: 
 * @Parm: 传入接受到的一个数据帧和长度    Pass in a received data frame and length
 * @Retval: 
 */
void Pto_Data_Parse(uint8_t *data_buf, uint8_t num)
{
    uint8_t pto_head = data_buf[0];
    uint8_t pto_tail = data_buf[num-1];
    
    //校验头尾  Check head and tail
    if (!(pto_head == PTO_HEAD && pto_tail == PTO_TAIL))
    {
        //DEBUG_PRINT("pto error:pto_head=0x%02x , pto_tail=0x%02x\n", pto_head, pto_tail);
        return;
    }
    
    uint8_t data_index = 1;
    uint8_t field_index[PTO_BUF_LEN_MAX] = {0};
    int i = 0;
    int values[PTO_BUF_LEN_MAX] = {0};
    char msgs[][PTO_BUF_LEN_MAX] = {0};
    
    //分割字段  Split Field
    for (i = 1; i < num-1; i++)
    {
        if (data_buf[i] == ',')
        {
            data_buf[i] = 0;
            field_index[data_index] = i;
            data_index++;
        }
    }
    
    //解析长度与功能ID  Parsing length and function ID
    for (i = 0; i < 2; i++)
    {
        values[i] = Pto_Char_To_Int((char*)data_buf+field_index[i]+1);
    }
    
    uint8_t pto_len = values[0];
    uint8_t pto_id = values[1];
​
    ParseCommonFields(pto_id,data_buf,pto_len,field_index,data_index,values,msgs);
​
}
​
​

Parses received data that conforms to the protocol, obtains the function ID and data information from the data, and executes the corresponding function handler. This project uses a metadata parsing method. The core idea is to separate the protocol format description (field positions, types, etc.) from parsing logic, defining protocol format through data structures (like struct arrays), and parsing functions automatically process data according to these descriptions.

Different types of data in the protocol can also be parsed. If you want to understand implementation details, you can study it yourself; no excessive explanation will be provided. This function has been adapted for all K230 communication protocols in our store, so it can be used normally without complete understanding.

• bsp_PID_motor.c

xxxxxxxxxx
​
​
// PID单独计算一条通道 PID calculates one channel separately
float PID_Calc_One_Motor(uint8_t motor_id, float now_speed)
{
    if (motor_id >= MAX_MOTOR)
        return 0;
​
        return PID_Location_Calc(&pid_motor[motor_id], now_speed);
        
        
}
​
void Set_PID_Motor(float set_l ,float set_r,float turn_out)
{
        
        
        l_pid_out = PID_Calc_One_Motor(0, set_l);
        r_pid_out = PID_Calc_One_Motor(1, set_r);
​
        PWM_Control_Car(l_pid_out+turn_out , r_pid_out-turn_out );
​
}
​
// 位置式PID计算方式
//Position PID calculation method
float PID_Location_Calc(PID_t *pid, float actual_val)
{
    /*计算目标值与实际值的误差*/
    /*Calculate the error between the target value and the actual value*/
    pid->err = pid->target_val - actual_val;
​
    /* 限定闭环死区 */
    /*Limited closed-loop dead zone*/
    if ((pid->err >= -40) && (pid->err <= 40))
    {
        pid->err = 0;
        pid->integral = 0;
    }
​
    /* 积分分离，偏差较大时去掉积分作用 */
    /*Integral separation, removing the integral effect when the deviation is large*/
    if (pid->err > -1500 && pid->err < 1500)
    {
        pid->integral += pid->err; // error accumulation 误差累积
​
        /* Limit the integration range to prevent integration saturation 限定积分范围，防止积分饱和 */
        if (pid->integral > 4000)
            pid->integral = 4000;
        else if (pid->integral < -4000)
            pid->integral = -4000;
    }
​
    /*PID算法实现*/ /*PID algorithm implementation*/
    pid->output_val = pid->Kp * pid->err +
                      pid->Ki * pid->integral +
                      pid->Kd * (pid->err - pid->err_last);
        
​
    /*误差传递*/ /*Error transmission*/
    pid->err_last = pid->err;
​
    /*返回当前实际值*/ /*Returns the current actual value*/
    return pid->output_val;
}
​

Set_PID_Motor: Sets target values for 2 motors and the turning loop output value.

PID_Calc_One_Motor: Calculates the position loop PID value for one channel and returns the calculated value.

PID_Location_Calc: Position loop PID calculation formula and complete position loop structure.

• bsp_K230_control.c

xxxxxxxxxx
void Target_Track(int x)
{
    
        PID_TypeDef k230_pid;
    PID_param_init(&k230_pid);
​
    set_p_i_d(&k230_pid, k230_p3, 0, k230_d3);
    int16_t err = 280 - x;
​
    if(err<5 && err>-5)err=0;
​
    pid_output = PID_Calculate(&k230_pid,err);
​
    if(pid_output>70) pid_output=70;
    if(pid_output<-70) pid_output=-70;
    
    
   Set_PID_Motor(0, 0, pid_output);
}
​
​
​

Target_Track: Calculates the offset of the recognized target on the x-axis and applies limiting to control the car's rotational tracking.

4. Experimental Operation

1. Connect hardware according to the tutorial's wiring instructions.
2. Open the project used by MSPM0G3507, then compile and burn the program, finally reset or power on again.
3. Place Target_Tracking.py in the root directory of K230's SD card and rename it to main.py, then reset or power on again. If you are not sure how to flash programs to K230, please first check the K230 Vision Module tutorial's [Quick Start] chapter.
4. Adjust the K230 module bracket to a suitable angle and connect the power supply.
5. During the first 6-7 seconds after successful K230 initialization, a green box will appear in the center of the screen. This is the learning process. Align this green box with the object you want to recognize (here we use an ultrasonic module as an example) and wait for about 8 seconds. When the box in the center of the screen turns red, it indicates entering the search and tracking phase.

Since this recognition function easily loses the target during rapid movement, the car only uses in-place rotational tracking. Forward-backward tracking will not be performed.

Note:

If you are not satisfied with the tracking effect, you can modify the PID values in the bsp_PID_motor.h file of the project.

5. Experimental Phenomena

Turn on the power switch, wait for system initialization to complete. The screen will display the camera feed, and a green box will appear in the center of the screen. After aligning this green box with the object you want to recognize and waiting for a period of time, when the box in the center of the screen turns red, the car will start rotational tracking in place.