Scribd
Upload
77 views12 pages

ECE 551 Lecture 7

Optimal Control Lecture Notes

Uploaded by

adambose1990
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
77 views12 pages

ECE 551 Lecture 7

Optimal Control Lecture Notes

Uploaded by

adambose1990
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 12
LecTURE 7 Consider the minimum-hme transfer ofe linear, stobroncsy system HG) = Axe) + buy =) Fromon arbitrary initial state 3, ot t2$ tosame target set SCL). The admissible controls satiety luce let e} The Hamiltonian i's. HC aC), ued, pC = 1+ pTAA nce) + ple) bult) ---® From the minimum principle we know thet if an optimal control exists it most sobs ty 1+ p TOA 2G) + p*Terbutay < 1+ phTCedA tee) + p* Tee) buted -@ for atl e [b, tp] and forall admissible ute). Now, since ty 1s free and the Hamiltonian dees net explicitly contamt, we have thet AC x8), whe, ph) = gb .® RCE 554 LECTURE 7 forall t € Ld, 4,1. From equation @), we observe that Ct, ,t,1 5 ao singelen interval fe pYEbLL=4 feralleelt 1] ..© This cannct happen because 1f we substitte Cer = bth equation®, then we heve a contra dicten, ve. 124. Thes, POYYS foronptelpe,] ..@ Equetion © 3 also sotiafied (forall t) if bed -.-® But, this indleotes thet the control dees not affect the system. So, let's consider the posei- bility thet the product p*"Ct)b = ¢ for bE Cet. 7. Pah tree, then duciig tht interval pee b = ¢ Sie LptTarbl=¢ fon ket aw ..@ Since bib an nxt mnateik of constants « ae See [ pte b] = SL ph tordb os, edb e .® Ece 551 LecTorRE 7 From the Hamiltonian , the covtote equotion is -@® aod tte selubion i x ae pase We where ¢ fs the vector of initial costote values. Let's enumerate some of the derivatives peo = 6 ple b= 6 pH b=¢ Bree =$ fortele,t,] ...@ Now, vewng equations @) ond (2) we hove thet pitere = Cate, hag -- @ By applying the matrix identity CMM, it - mtMT we get -Ae oT Ce WAab=¢ ---@ Lecture 9 Similarly, diPlerentiating equation @® gives Ty = - a" pte mG St for plwreg Switehing® cceur when ple) = 6. IF Shere d= die interval L4, I due “3 winch perf forallte Cet, 1 -. then Lt, 4, 1.8 aeingulon intecval. S Free ond df si not explicitly dependent on then we hove thet ECE 551 LECTURE 7 BEG) 227A) + pitted ace + pheentun = forall tele, 1. Mf preys fortel then pecedepste) # plGss eg eet, t, 1. Also, from equeten@l) we have thet MES Mart 2 ter epralgiar=¢ ...@D forte ltt, 1, and hence eWea ces Ce 4. MeM=M gp, tele,4,3 Tfe einguler intervol is to exist. The costate equetions are PrGy= = xb) po Ged = — af) - pte) Dor aga singuler interval, wang eave and @) we cbtern prey = -a2@) Substituting thi inte equetion B® yields x2) xP gS ...@ 10 Ece 551 LecTURE 7 "1 Latare tar ixtar-gted=d ... Note thet cqueton GD is sotistied if HPQ) + xhLr= gp .. ® x8) ~ x)! =¢ ---@ for t6Lt,,t,]. Ewe differentiate equation @) ard usethe result te substitute vm equat n@© we get BPCLY * HAY © SL) ---@) tlh) for tele, 4,7 -.@ @®Dend use the result sn the stote equation we get locly, if we differentiote equate wr) = ate) forte lt,t,7 Thos, equehiers @ and @) delice alocus of ports inthe state plone where cinguler co. @ + Dons @ are explicit expressions for the evagvler control law. The, meg exit. Equa singulec lines , troncoted a Ix curl = I since, Bce 551 Lecture 7 12 fuer] 41, ave shown below. ma) The Line myray, carrot be pert ofan opt mel 2. trejectory , since along this ine the system moves awey from the o

You might also like