Skip to main content

Table 2 Parameters and symbols of MPC problems

From: A Combined Reinforcement Learning and Model Predictive Control for Car-Following Maneuver of Autonomous Vehicles

Parameters

Symbols

\(t\)

Current moment

\(N_{p}\)

Prediction horizon

\(N_{c}\)

Control horizon

\(J_{t}\)

The objective function of the longitudinal motion planning

\(x(0)\)

The state vector at the current moment

\(u_{t - 1}\)

The control vector at the previous moment

\(y_{{t + it_{p} \left| t \right.}}\)

The predictive output of longitudinal position and velocity corresponding to each predictive step in the prediction horizon at the current step

\(y_{{{\text{ref}},t + it_{p} \left| t \right.}}\)

The desired output of longitudinal position and velocity of each predictive step in the prediction horizon

\(u_{{t + jt_{c} \left| t \right.}}\)

The output of acceleration corresponding to each predictive step in the control horizon at the current step

\({\Delta }u_{{t + it_{c} \left| t \right.}}\)

The output of jerk corresponding to each predictive step in the control horizon at the current step

\(\varepsilon\)

Relaxation factor

\(Q,R_{u} ,R_{du} ,\rho\)

Weights of each optimization objectives

\(1_{p \times 1} ,1_{n \times 1}\)

A unit column vector of dimension n and p