python模型预测控制_模型预测控制

【实例简介】

About

This work aims at implementing simple MPC controller for gym's Mujoco models as described in Neural Network Dynamics for Model-Based Deep Reinforcement Learning with Model-Free Fine-Tuning and build on it by adding LQR based controllers instead of using simple shooting methods. Such controllers are then applied in parallel and the stored trajectories are used to learn a general neural network policy.

Dependencies

This code has been tested on python3 and requires mujoco_py installed.

How to Run

Please use python3 main.py to run. Passing --load_model will restore the previously stored policy parameters.

【实例截图】

【核心代码】

MPC-master

├── README.md

├── __pycache__

│   ├── cheetah_env.cpython-35.pyc

│   ├── controllers.cpython-35.pyc

│   ├── cost_functions.cpython-35.pyc

│   ├── dynamics.cpython-35.pyc

│   └── logz.cpython-35.pyc

├── cheetah_env.py

├── controllers.py

├── cost_functions.py

├── data

│   ├── mb_mpc_HalfCheetah-v1_10-01-2018_19-26-31

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_10-01-2018_19-27-37

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_10-01-2018_19-28-13

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_11-16-05

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_11-18-18

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_11-21-23

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_11-21-56

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_11-43-58

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_11-51-29

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_11-51-45

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_11-51-58

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_11-54-06

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_12-00-39

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_12-01-58

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_12-17-09

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_12-18-25

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_12-30-29

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_12-31-54

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_12-32-14

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_12-33-12

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_12-34-21

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_13-04-22

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_13-04-49

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_13-05-08

│   │   └── log.txt

│   ├── mb_mpc_HalfCheetah-v1_11-01-2018_13-05-24

│   │   └── log.txt

│   └── mb_mpc_HalfCheetah-v1_11-01-2018_13-06-02

│       └── log.txt

├── dynamics.py

├── iLQR.py

├── logz.py

├── main.py

├── plot.py

└── policy.py

28 directories, 41 files

你可能感兴趣的:(python模型预测控制)