- #1
ra_forever8
- 129
- 0
Investigate the stability of the PECE method where
P=Predictor : y_(n+1) = y_n + hf(y_n)
C=Corrector: y_(n+1) = y_n + h [(1-θ) f(y_n) + θ f(y_(n+1))], (0<θ<1)
and E is the evaluation step.
=>
substituting the predictor into corrector gives:y_(n+1) = y_n + h [(1-θ) f(y_n) + θf( y_n+ h f y_n )]...(1)
stability : dy/dt = λy=f
which has exact solution of y= yo e^(λt)
substitute f=λy into (1)
y_(n+1) = y_n + h [(1-θ) λ y_n + θλ( y_n+ h λy_n )]
y_(n+1) = y_n + h ( λ y_n-θ λ y_n + θλ y_n+ θh λ^2 y_n )
y_(n+1) = y_n + h ( λ y_n+ θh λ^2 y_n )
y_(n+1) = y_n + h λ y_n+ θ(h λ)^2 y_n
y_(n+1) = y_n (1+ h λ+ θ(hλ)^2)
now,
|1+ h λ+ θ(hλ)^2|<1
-1 < 1+ h λ+ θ(hλ)^2 < 1
Kindly anyone help me after this , how to investigate the stability of PECE method
P=Predictor : y_(n+1) = y_n + hf(y_n)
C=Corrector: y_(n+1) = y_n + h [(1-θ) f(y_n) + θ f(y_(n+1))], (0<θ<1)
and E is the evaluation step.
=>
substituting the predictor into corrector gives:y_(n+1) = y_n + h [(1-θ) f(y_n) + θf( y_n+ h f y_n )]...(1)
stability : dy/dt = λy=f
which has exact solution of y= yo e^(λt)
substitute f=λy into (1)
y_(n+1) = y_n + h [(1-θ) λ y_n + θλ( y_n+ h λy_n )]
y_(n+1) = y_n + h ( λ y_n-θ λ y_n + θλ y_n+ θh λ^2 y_n )
y_(n+1) = y_n + h ( λ y_n+ θh λ^2 y_n )
y_(n+1) = y_n + h λ y_n+ θ(h λ)^2 y_n
y_(n+1) = y_n (1+ h λ+ θ(hλ)^2)
now,
|1+ h λ+ θ(hλ)^2|<1
-1 < 1+ h λ+ θ(hλ)^2 < 1
Kindly anyone help me after this , how to investigate the stability of PECE method