Control engineering — 学习指南
Distinguish feedback, estimation and optimization, and test responses in PID and Kalman experiments.
按照清晰的学习路径，从基础原理走向实际应用。
Before you start Basic functions, calculus and vectors are used. Topic prerequisites link to state-space and closed-loop explanations.
1. Feedback control
Read how PID works, then vary delay and saturation.
-
PID control
Background: Closed-loop feedback
- Primer控制工程入门：PID控制——跨三个时间尺度的工作
- Lab · Run in browser在Python中尝试PID控制——增益、延迟和饱和度
2. Estimating state from observations
Separate prediction and update, then vary noise assumptions.
-
Kalman filtering
Background: State-space model
- Primer控制工程入门：状态观测器和线性卡尔曼滤波器
- Lab · Run in browser在 Python 中运行卡尔曼滤波器：噪声、缺失数据和异常值
3. Choosing inputs with costs and constraints
Move from LQR weights to MPC constraints and repeated optimization.
-
LQR
Background: State-space model
- Primer控制工程入门：LQR——基于显式优先级的状态反馈
-
Model predictive control
Background: Closed-loop feedback · State-space model
- Primer控制工程入门：模型预测控制——在约束条件下运行
4. Timing a motion
Distinguish path geometry from velocity and acceleration timing.
-
Trajectory timing
5. Following a path
Move from geometric lookahead tracking to speed regulation and front-axle error steering on one shared vehicle model.
-
Pure Pursuit and Adaptive Pure Pursuit
Background: Trajectory timing
- Lab · Run in browser纯追踪实验室——比较固定前视距离的路径跟踪
- Lab · Run in browser自适应纯追踪实验室——让前视距离随速度缩放
-
Regulated Pure Pursuit speed regulation
Background: Pure Pursuit and Adaptive Pure Pursuit
- Lab · Run in browser受调纯追踪实验室——弯道和接近目标时的速度
-
Stanley control
Background: Pure Pursuit and Adaptive Pure Pursuit
- Lab · Run in browserStanley 实验室——由前轴横向跟踪误差和航向误差决定转向
-
Path tracking with MPC
Background: Stanley control · Model predictive control
- Lab · Run in browserMPC 实验室——在时域和转向限制内重新求解曲率序列
-
Comparing path trackers
Background: Regulated Pure Pursuit speed regulation · Path tracking with MPC
- Lab · Run in browser路径跟踪比较实验室——在相同条件下运行 PP、APP、RPP、Stanley 和 MPC