1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17 """Computes 2nd derivatives of complex power flow w.r.t. voltage.
18 """
19
20 from numpy import ones, conj
21 from scipy.sparse import csr_matrix
22
23
25 """Computes 2nd derivatives of complex power flow w.r.t. voltage.
26
27 Returns 4 matrices containing the partial derivatives w.r.t. voltage angle
28 and magnitude of the product of a vector C{lam} with the 1st partial
29 derivatives of the complex branch power flows. Takes sparse connection
30 matrix C{Cbr}, sparse branch admittance matrix C{Ybr}, voltage vector C{V}
31 and C{nl x 1} vector of multipliers C{lam}. Output matrices are sparse.
32
33 For more details on the derivations behind the derivative code used
34 in PYPOWER information, see:
35
36 [TN2] R. D. Zimmerman, I{"AC Power Flows, Generalized OPF Costs and
37 their Derivatives using Complex Matrix Notation"}, MATPOWER
38 Technical Note 2, February 2010.
39 U{http://www.pserc.cornell.edu/matpower/TN2-OPF-Derivatives.pdf}
40
41 @author: Ray Zimmerman (PSERC Cornell)
42 @author: Richard Lincoln
43 """
44 nb = len(V)
45 nl = len(lam)
46 ib = range(nb)
47 il = range(nl)
48
49 diaglam = csr_matrix((lam, (il, il)))
50 diagV = csr_matrix((V, (ib, ib)))
51
52 A = Ybr.H * diaglam * Cbr
53 B = conj(diagV) * A * diagV
54 D = csr_matrix( ((A * V) * conj(V), (ib, ib)) )
55 E = csr_matrix( ((A.T * conj(V) * V), (ib, ib)) )
56 F = B + B.T
57 G = csr_matrix((ones(nb) / abs(V), (ib, ib)))
58
59 Haa = F - D - E
60 Hva = 1j * G * (B - B.T - D + E)
61 Hav = Hva.T
62 Hvv = G * F * G
63
64 return Haa, Hav, Hva, Hvv
65