Custom Reinforcement Environment and Building Agent H

Question

I'm trying to build a RL environment and an agent, and I'm having some difficulties to understand things. First of all, my problem is to fit 2 curves by some rules. I think I managed to create a custom enviroment but I couldn't figure out how to build an agent. In the first figure you can see I have 2 curves. Second curve will be stable and first curve will be fit on it at some rules. My curves as an example

This is my environment.

class testEnv(Env):
    def __init__(self):
        # Axtions we can take: up, down, wait
        # self.action_space = Discrete(3)
        # Action array
        self.action_space = Box(low=-1.0,high=1.0,dtype=np.float32)

        self.observation_space = Box(low=np.array([-100]),high = np.array([100]),dtype=np.float32)
        #Set start amp array

        self.state = x_rl1 + random.uniform(-0.5,0.5)
        # Set time (60 sec)
        self.time_length = 60

    def step(self,action):

        self.state += action - 0.1

        self.time_length -= 1

        TMS_Env = 0.67
        y_Env =[]

        for x in self.state:
            y_Env.append((TMS_Env * ((0.14)/(x**0.02-1))))

        y_Env = np.array(y_Env)

        dt = np.min(np.subtract(self.state,y_Env))

        if dt<0.4 or dt>0.29:
            if dt == 0.4:
                reward = 300
                done = True
            else:
                reward = 1
        else:
            reward = -1

        if self.time_length <= 0:
            done = True
        else:
            done = False
        # Noise
        self.state += random.uniform(-1,1)
        info = {'dt:{}'.format(dt),'y_Env:{}'.format(y_Env),'self.state:{}'.format(self.state)}

        return self.state, reward, done, info,y_Env

    def reset(self):
        self.state = np.linspace(1.1,30,30) #+ random.randint(-2,2)
        self.time_length = 80
        pass

    def render(self):
        pass

When I run this manuelly I can get this:

enter image description here

What I want to do is, like I said it earlier, build an agent. What I know is DQN won't work for me because my action space is in BOX type. So, I decided to use DDPG. At this point, I stuck. I don't know what I'm going to do now.