问题描述
设连续系统状态方程和性能指标
X
˙
=
f
(
t
,
X
,
U
)
X
(
t
0
)
=
X
0
J
=
ϕ
[
X
(
t
f
)
,
t
f
]
+
∫
t
0
t
f
F
(
X
,
U
,
t
)
d
t
\begin{aligned} \dot{X} & =f(t, X, U) \quad X\left(t_{0}\right)=X_{0} \\ J & =\phi\left[X\left(t_{f}\right), t_{f}\right]+\int_{t_{0}}^{t_{f}} F(X, U, t) d t \end{aligned}
X˙J=f(t,X,U)X(t0)=X0=ϕ[X(tf),tf]+∫t0tfF(X,U,t)dt
HJI与HJB方程
HJB方程
设定如下哈密顿函数定义:
H
(
X
,
U
,
λ
,
t
)
=
F
(
X
,
U
,
t
)
+
λ
T
f
(
X
,
U
,
t
)
λ
=
∂
V
∂
X
\begin{aligned} H(X, U, \lambda, t)=&F(X, U, t)+\lambda^{T} f(X, U, t) \\ \lambda=&\frac{\partial V}{\partial X} \end{aligned}
H(X,U,λ,t)=λ=F(X,U,t)+λTf(X,U,t)∂X∂V
HJB方程
−
∂
V
∂
t
=
min
u
∈
Ω
H
(
X
,
U
,
∂
V
∂
X
,
t
)
=
H
∗
(
X
,
U
,
∂
V
∂
X
,
t
)
-\frac{\partial V}{\partial t}=\min _{u \in \Omega} H\left(X, U, \frac{\partial V}{\partial X}, t\right)=H^{*}\left(X, U, \frac{\partial V}{\partial X}, t\right)
−∂t∂V=u∈ΩminH(X,U,∂X∂V,t)=H∗(X,U,∂X∂V,t)
此外,参数
λ
\lambda
λ满足协态方程
λ
˙
=
−
∂
H
∂
X
\dot{\lambda}=-\frac{\partial H}{\partial X}
λ˙=−∂X∂H
横截条件
λ
(
t
f
)
=
∂
ϕ
∂
X
(
t
f
)
\lambda(t_f)=\frac{\partial \phi}{\partial X(t_f)}
λ(tf)=∂X(tf)∂ϕ
HJI方程
考虑一个博弈问题其价值函数为
V
(
x
)
=
min
u
p
max
u
e
J
V(\boldsymbol{x})=\min _{\boldsymbol{u}_{p}} \max _{\boldsymbol{u}_{e}} J
V(x)=upminuemaxJ
HJI方程
−
∂
V
∂
t
=
∂
V
∂
x
f
(
x
)
+
F
(
x
)
-\frac{\partial V}{\partial t}=\frac{\partial V}{\partial \boldsymbol{x}}f(x)+F(x)
−∂t∂V=∂x∂Vf(x)+F(x)