ros读取cmd_vel话题数据并输出到图片

from envforgym import SimEnv
from stable_baselines3 import PPO
from geometry_msgs.msg import Twist
from torch.utils.tensorboard import SummaryWriter
import rospy
import matplotlib.pyplot as plt

class Test():
    def __init__(self):
        self.linear_velocity = 0
        self.angular_velocity = 0
        self.linear = []
        self.angular = []
    def control_callback(self, data):
        self.linear_velocity = data.linear.x
        self.angular_velocity = data.angular.z
        if self.linear_velocity > 0:
            self.linear.append(self.linear_velocity)
            self.angular.append(self.angular_velocity)


if __name__ == '__main__':
    env = SimEnv()
    test = Test()
    agent = PPO.load('/home/houjj/RLnavigation/src/models/PPO_2/368640',env=env)

    obs = env.reset()[0]
    terminated = False
    step = 0

    odom_receiver = rospy.Subscriber("/cmd_vel", Twist, test.control_callback, queue_size=1)

    while not terminated:
        step += 1
        action,_state = agent.predict(obs)
        obs,rew,terminated,truncated,info = env.step(action)

    plt.figure()
    plt.subplot(2, 1, 1)
    plt.plot(test.linear)
    plt.title('linear_velocity')
    plt.subplot(2, 1, 2)
    plt.plot(test.angular)
    plt.title('angular_velocity')
    plt.show()

上面这个是可行的，下面这个却不可行，不知道为什么，记录以下

from envforgym import SimEnv
from stable_baselines3 import PPO
from geometry_msgs.msg import Twist
from torch.utils.tensorboard import SummaryWriter
import rospy

def control_callback(data):
    global linear_velocity
    global angular_velocity
    linear_velocity = data.linear.x
    angular_velocity = data.angular.z
    pass

if __name__ == '__main__':
    env = SimEnv()
    agent = PPO.load('/home/houjj/RLnavigation/src/models/PPO/307200',env=env)
    log_dir = 'testlog'

    writer = SummaryWriter(log_dir=log_dir)

    obs = env.reset()[0]
    terminated = False
    step = 0

    linear_velocity = 0
    angular_velocity = 0

    odom_receiver = rospy.Subscriber("/cmd_vel", Twist, control_callback, queue_size=1)

    while not terminated:
        step += 1
        action,_state = agent.predict(obs)
        obs,rew,terminated,truncated,info = env.step(action)
        print("linear_velocity:", linear_velocity, "angular_velocity:", angular_velocity)


        writer.add_scalar("linear_velocity", linear_velocity, step)
        writer.add_scalar("angular_velocity", angular_velocity, step)

    # 关闭writer
    writer.close()

    print("done")

    # 这里缺少了ROS的主循环，需要添加ros.spin()来处理消息
    rospy.spin()