Steady State Errors ELEC304-Alper Erdogan 1–1 Lecture 1 Steady State Errors • We define the error as the difference between the input and output: • For unity feedback systems error is the input of the feedforward block: And the steady state error e(∞) = lim sE(s) s→0 (1) Steady State Errors ELEC304-Alper Erdogan 1–2 Steady State Error in Terms of T (s) • Error can be written as E(s) = R(s) − C(s) Here C(s) = R(s)T (s) and therefore E(s) = R(s) − R(s)T (s) = R(s)(1 − T (s)) As a result, the steady state error e(∞) = lim sR(s)(1 − T (s)) s→0 Example: For T (s) = R(s) = 1s 5 s2 +7s+10 and unit step input 1 s2 + 7s + 5 E(s) = (1 − T (s)) = s s(s2 + 7s + 10) Steady State Errors ELEC304-Alper Erdogan and the steady state error s2 + 7s + 5 1 e(∞) = lim sE(s) = lim 2 = s→0 s→0 s + 7s + 10 2 How about the steady state error for the ramp input (R(s) = s12 )? 1 s2 + 7s + 5 E(s) = 2 (1 − T (s)) = 2 2 s s (s + 7s + 10) and the steady state error s2 + 7s + 5 e(∞) = lim sE(s) = lim =∞ s→0 s→0 s(s2 + 7s + 10) 1–3 Steady State Errors ELEC304-Alper Erdogan 1–4 Steady State Error in Terms of G(s) • Note that E(s) = R(s) − C(s) and C(s) = E(s)G(s) which yield E(s) = R(s) 1 + G(s) Therefore the steady state error sR(s) e(∞) = lim sE(s) = lim s→0 s→0 1 + G(s) • For a step input R(s) = 1 s s 1s 1 e(∞) = lim = s→0 1 + G(s) 1 + lims→0 G(s) Steady State Errors ELEC304-Alper Erdogan 1–5 To have zero steady state error lim G(s) = ∞ s→0 which implies G(s) must take the form (s + z1)(s + z2)... G(s) = n n≥1 s (s + p1)(s + p2)... i.e. at least one pure integrator! • For a ramp input R(s) = 1 s2 1 s s12 = e(∞) = lim s→0 1 + G(s) lims→0 sG(s) To have zero steady state error lim sG(s) = ∞ s→0 which implies G(s) must take the form (s + z1)(s + z2)... n≥2 G(s) = n s (s + p1)(s + p2)... i.e. at least two pure integrators! • For a parabolic input R(s) = 1 s3 s s13 1 e(∞) = lim = s→0 1 + G(s) lims→0 s2G(s) To have zero steady state error lim s2G(s) = ∞ s→0 Steady State Errors ELEC304-Alper Erdogan 1–6 which implies G(s) must take the form G(s) = (s + z1)(s + z2)... sn(s + p1)(s + p2)... i.e. at least three pure integrators! n≥3 Steady State Errors ELEC304-Alper Erdogan 1–7 Examples Consider the System (with no integrator) in the following figure • The steady state error for r(t) = 5u(t) (R(s) = 5s ) s 5s 5 5 5 e(∞) = lim = = = s→0 1 + G(s) 1 + lims→0 G(s) 1 + 20 21 • The steady state error for r(t) = 5tu(t) (R(s) = s s52 5 e(∞) = lim = =∞ s→0 1 + G(s) lims→0 sG(s) 5 ) s2 Steady State Errors ELEC304-Alper Erdogan 1–8 Examples Consider the System (with one integrator) in the following figure • The steady state error for r(t) = 5u(t) (R(s) = 5s ) s 5s 5 5 e(∞) = lim = = =0 s→0 1 + G(s) 1 + lims→0 G(s) ∞ • The steady state error for r(t) = 5tu(t) (R(s) = 5 ) s2 5 5 1 s s52 = = = e(∞) = lim s→0 1 + G(s) lims→0 sG(s) 100 20 • The steady state error for r(t) = 5t2u(t) (R(s) = 10 ) s3 s 10 10 10 s3 e(∞) = lim = = =∞ 2 s→0 1 + G(s) lims→0 s G(s) 0 Steady State Errors ELEC304-Alper Erdogan 1–9 Type of the System The type of the system (the number of integrators in the open loop system) determines which type of signals we can track: Here n is the type of the system. • If n = 0, – It can track unit step with a nonzero error 1 1 e(∞) = = 1 + lim G(s) 1 + Kp s→0 {z Kp | } where Kp is the position constant. – It can not track ramp or parabolic inputs. • If n = 1, – It can track unit step with a zero error. – It can track ramp with a nonzero error. 1 1 e(∞) = = lim sG(s) Kv s→0 | {z Kv } Steady State Errors ELEC304-Alper Erdogan 1 – 10 where Kv is the velocity constant. – It can not track parabolic inputs. • If n = 2, – It can track unit step and ramp input with zero errors. – It can track parabolic input with a nonzero error. 1 1 = e(∞) = lim s2G(s) Ka s→0 | {z } Ka where Ka is the acceleration constant. Steady State Errors ELEC304-Alper Erdogan 1 – 11 Steady State Errors: Static Error Constants • Position constant (1) Kp = lim G(s) s→0 Steady state error for unit step input is e(∞) = 1 1+Kp . • Velocity constant Kv = lim sG(s) (2) s→0 Steady state error for a ramp input is e(∞) = 1 KV . • Acceleration constant Ka = lim s2G(s) s→0 Steady state error for a parabolic input is e(∞) = (3) 1 Ka . Steady State Errors ELEC304-Alper Erdogan 1 – 12 Example Given the control system find K such that there is a 10 percent error in steady state? Solution: Since system type is 1 the finite steady state error should be for a ramp input. e(∞) = So 1 = 0.1 Kv 5·K Kv = 10 = lim sG(s) = s→0 6·7·8 we obtain K = 672 Steady State Errors ELEC304-Alper Erdogan How About Nonunity Feedback Systems • Easy! Use the following transformation: 1 – 13