hyCPN1
hybrid CP decomposed norm-1 tensor for dmss model parameters
Contents
Syntax
H = hyCPN1() Create empty hyCPN1 tensor with hybrid parameter structure. For information on how to access and change model parameters look into examples.
Description
hyCPN() stores the model paramters of a dmss model in the structure of a hybrid norm-1 CP decomposed tensor. The norm-1 CP-decomposed tensor format has been introduced in [2]. A hybrid implicit multilinear time invariant (hybrid iMTI) model has been briefly introduced in [1].
hybrid structure:
For the structure matrix H.F, the rows of H.F have been split into the different variable types:
H.F.state = [];
H.F.stateDerivative = [];
H.F.algebraic = [];
H.F.boolean = []; --> Experimental: For hybrid dmss models
H.F.input = [];
For the parameter matrix H.phi, the rows of H.phi have been split into the different equations and inequality constraints (for hybrid dmss models):
H.phi.equality = [];
H.phi.inequality = []; --> Experimental: For hybrid dmss models
Matrix structure:
Matrices in these structures will stored in a hybrid fashion with the following sub structure as
For the structure matrix H.F:
-> H.F.state.t stores all true*x (equivalent to 1*x) values of H.F.state as sparse logical matrix
-> H.F.state.f stores all false*x (equivalent to (1-x)) values of H.F.state as sparse logical matrix
-> H.F.state.c stores all continuous valued parameters (equivalent to (1-F + F*x)) of H.F.state as sparse double/symbolic matrix
For the parameter matrix H.phi:
-> H.phi.equality.t stores all true (equivalent to 1) values of H.F.state as sparse logical matrix
-> H.F.state.f stores all false (equivalent to -1) values of H.F.state as sparse logical matrix
-> H.F.state.c stores all continuous valued parameters of H.F.state as sparse double/symbolic matrix
Examples
To generate a model using an hyCPN1 object, first an empty hyCPN1 object must be created
H = hyCPN1()
H = hyCPN1 with properties: F: [1×1 struct] phi: [1×1 struct]
Now the structure of H can be filled with matrices by accessing its properties. It can be either filled by logical, double or symbolic values.
H.F.stateDerivative = [1 0 0 0 0 0 0 0]; H.F.state = [0 1 0 0 1 0 sym("k_3") sym("k_5")]; H.F.input = [0 1 1 0 0 0 sym("k_2") sym("k_4");... 0 0 1 0 0 0 0 0;... 0 0 0 0 0 1 0 0]; H.F.algebraic = [0 0 0 1 0 0 0 0;... 0 0 0 0 1 1 sym("k_1") 0]; H.phi.equality = [sym("m_w")*sym("c_w") sym("c_w") -sym("c_w") 1 0 0 0 0;... 0 0 0 1 -sym("A") sym("A") 0 0;... 0 0 0 0 0 0 sym("phi_1") sym("phi_2")];
When setting a sub matrix (e.g. H.F.state) the matrix will be stored in a sparse hybrid fashion (H.F.state.t, H.F.state.f and H.F.state.c), for making use of faster sub-routines in simulation etc. Moreover these metrices are used in different routines, like in connecting different dmss models witch connect(). Also those sub matrices can be accessed, set or changed directly, e.g. via H.F.state.t = [0 1 0 0 1 0 0 0].
Now a dmss object can be created from the filled hyCPN1 object.
sys = dmss(H)
sys = dmss with properties: H: [1×1 hyCPN1] n: 1 m: 3 p: 2 outputIndex: [1×1 struct] nEq: 3 stateName: [0×0 string] stateUnit: [0×0 string] algebraicName: [0×0 string] algebraicUnit: [0×0 string] inputName: [0×0 string] inputUnit: [0×0 string] ts: 0
Symbolic equations can be generated with
disp(sys.symbolicEquations)
y1 + c_w*m_w*xp1 - c_w*u1*u2 + c_w*u1*x1 == 0 y1 + A*u3*y2 - A*x1*y2 == 0 phi_2*(u1*conj(k_4) - abs(conj(k_4)) + 1)*(x1*conj(k_5) - abs(conj(k_5)) + 1) + phi_1*(u1*conj(k_2) - abs(conj(k_2)) + 1)*(x1*conj(k_3) - abs(conj(k_3)) + 1)*(y2*conj(k_1) - abs(conj(k_1)) + 1) == 0
e.g. for checking the correctness model. Symbolic parameters can be changed by
symParameters = [sym("m_w") sym("c_w") sym("A") sym("phi_1") sym("phi_2") sym("k_1") sym("k_2") sym("k_3") sym("k_4") sym("k_5")]; di = 206.5/1000; A = di^2 *pi()/4; m_w = A*998*1; c_w = 4200; k = [-0.1497, -2.3659, -0.0109, 0.5556, 0.0131, 0.9872, 0.0259]; numParameters = [m_w, c_w, A, k]; sys = sys.replaceSymbolicParameters(symParameters, numParameters); symParameters = [sym("m_w") sym("c_w") sym("A") sym("phi_1") sym("phi_2") sym("k_1") sym("k_2") sym("k_3") sym("k_4") sym("k_5")]; disp(sys.symbolicEquations)
(2411735508342479*xp1)/17179869184 + y1 - 4200*u1*u2 + 4200*u1*x1 == 0 y1 + (4826582636502955*u3*y2)/144115188075855872 - (4826582636502955*x1*y2)/144115188075855872 == 0 (1497*((1389*u1)/2500 + 1111/2500)*((131*x1)/10000 + 9869/10000)*((109*y2)/10000 - 9891/10000))/10000 - (23659*((617*u1)/625 + 8/625)*((259*x1)/10000 + 9741/10000))/10000 == 0
References
[1] G. Lichtenberg et al., "Implicit multilinear modeling," at - Automatisierungstechnik, vol. 70, no. 1, pp. 13–30, Jan. 2022. doi:10.1515/auto-2021-0133
[2] N. Jöres et al., “Reduced CP representation of multilinear models,” Proceedings of the 12th International Conference on Simulation and Modeling Methodologies, Technologies and Applications, 2022. doi:10.5220/0011273100003274
See Also
dmss, sym2dmss, mss2dmss, connect, replaceSymbolicParameters
Author(s): Torben Warnecke