ROS learning record ⑥: dynamic configuration parameters

1. Create a new feature package

cd catkin_ws/src
catkin_create_pkg pid roscpp rospy rosmsg std_msgs dynamic_reconfigure 

2. Create cfg file

Create a new cfg folder and create a new .cfgfile in it

mkdir cfg
cd cfg
touch PID.cfg

The core API of dynamic parameter tuning:

gen.add(name,type, level,description, default, min,max)

Parameter Description:

name: variable name
type: type name, commonly used types are: int_t, double_t, str_t, bool_t
level: a marker bit, as long as the parameter is modified, it will be changed to this value
description: parameter description
default: default value
min: available Optional, the minimum value of the parameter
max: optional, the maximum value of the parameter

For example, the content of dynamically adjusting PID parameters is as follows:

#! /usr/bin/env python
# coding=utf-8

import roslib
from dynamic_reconfigure.parameter_generator_catkin import *

PACKAGE = "pid"
roslib.load_manifest(PACKAGE)

# 创建一个参数生成器
gen = ParameterGenerator()

# 添加参数说明，便于后续生成界面
#             参数名    类型               等级       参数描述             默认值        最小值         最大值
gen.add("p",		double_t,		0,		"KP param.",			0.0,			-100.00,		100.00)
gen.add("i",		double_t,		0,		"KI param.",			0.0,			-100.00,		100.00)
gen.add("d",		double_t,		0,		"KD param.",			0.0,			-100.00,		100.00)

# 调用生成器生成config配置文件
#                                   包名           节点名称    生成文件名
exit(gen.generate(PACKAGE, PACKAGE, "PID"))

After that, set PID.cfg as an executable file, you can right-click and set properties, or set it through the command line:

chmod a+x PID.cfg

Also turn on the following comments in CMakelist.txt:

## Generate dynamic reconfigure parameters in the 'cfg' folder
generate_dynamic_reconfigure_options(
	cfg/PID.cfg		# 注意这里要填自己的路径
)

3. Node file

The steps of creating a new py node file are the same as normal, so I won’t go into details here

The case here writes a PID dynamic tuning parameter, here is the direct code and comments:

# 导入依赖
from pid.cfg import PIDConfig
from dynamic_reconfigure.server import Server
from std_msgs.msg import Float32MultiArray
import rospy


class UpdatePID():

    def __init__(self):
        rospy.init_node("update_pid")
        rospy.on_shutdown(self.shutdown)
        
        # 执行频率
        rate = rospy.Rate(20)
        
        # 声明一个消息发布者, 将消息发布到driver
        self.publisher_pid = rospy.Publisher("/pid", Float32MultiArray, queue_size=100)
        
        # 启动参数配置服务器         当参数变化时的回调
        Server(PIDConfig, self.dynamic_callback)
        
        # 打印一个启动成功的日志
        rospy.loginfo("updatePID_node start success! Bring up rqt_reconfigure to control the pid ...")
        
        # 让ros节点跑起来
        while not rospy.is_shutdown():
            rate.sleep()
            
	# 通过回调函数中获取rqt界面上的数值
    def dynamic_callback(self, config, level):
        """
        当参数发生变化时的回调函数
        :param config: 配置文件
        :param level: 修改的状态
        :return:
        """
        # 封装消息
        kp = config["p"]
        ki = config["i"]
        kd = config["d"]

        pid_msg = Float32MultiArray()
        pid_msg.data.append(kp)
        pid_msg.data.append(ki)
        pid_msg.data.append(kd)

        # 将消息发布出去
        self.publisher_pid.publish(pid_msg)

        print (config, level)
        return config

    def shutdown(self):
        print ("stop robot ...")


if __name__ == '__main__':
    UpdatePID()

After that, remember to give the py file executable permissions

chmod a+x updatePID_node.py

4. Start configuration

runroscore

roscore

catkin_makeone time

cd ~/catkin_ws
catkin_make

start node file

source devel/setup.bash
rosrun pid UpdatePIDNode.py 

At this point, remember to open rqt_reconfigurethe dynamic parameter adjustment

rosrun rqt_reconfigure rqt_reconfigure

At this time, you can also rqt_topiccheck whether the node is running normally in

Then send the pid parameters to the lower computer according to actual needs to realize dynamic parameter adjustment

---

• This blog post has been synchronously updated to Gu Yueju, link: ROS Learning Record – Dynamic Parameter Adjustment