EngneeringOptimization/SolverMethod.cpp at master · trunone/EngneeringOptimization · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
#include "BasicOptimizer.h"

CBasicOptimizer::CSolverMethod::CSolverMethod(CBasicOptimizer *p) : Super(p),
    FinalRange(0.00001) {
    n = Super->nDim;
}

void CBasicOptimizer::CSolverMethod::SteepestDecentMethod() {
    for(;;) {
        DifferentialMethod();

        if(norm() < FinalRange) break; // if arrive stop condition then end

        LineSearchMethod();
    }
}

void CBasicOptimizer::CSolverMethod::ConjugateGradientMethod() {
    int k = 0;

    for(;;) {

        if (k == 0) {
            DifferentialMethod();

            if(norm() < FinalRange) break; // if arrive stop condition then end
        }

        pfX_past = (double*)memcpy(
                       new double[Super->nDim],
                       Super->pfX,
                       Super->nDim*sizeof(double)
                   );

        LineSearchMethod();

        f_g_past = (double*)memcpy(
                       new double[Super->nDim],
                       Super->f_g,
                       Super->nDim*sizeof(double)
                   );

        DifferentialMethod();

        if(norm() < FinalRange) break; // if arrive stop condition then end

        double beta;

        switch(Super->CFormula) {
        case HestenesStiefel:
            beta = HestenesStiefelFormula();
            break;
        case PolakRibiere:
            beta = PolakRibiereFormula();
            break;
        case FletcherReeves:
            beta = FletcherReevesFormula();
            break;
        case Powell:
            beta = PowellFormula();
            break;
        default:
            break;
        }

        for(int i = 0; i<Super->nDim; i++)
            Super->f_g[i] += beta * f_g_past[i];

        if (k<=n) k++;
        else {
            Super->pfX = (double*)memcpy(
                             new double[Super->nDim],
                             pfX_past,
                             Super->nDim*sizeof(double)
                         );
            k = 0;
        }
        delete [] f_g_past;
        delete [] pfX_past;
    }
}

void CBasicOptimizer::CSolverMethod::DifferentialMethod() {
    CDifferentialMethod *DifferentialMethod = new CDifferentialMethod(Super);
    delete Super->f_g;
    switch(Super->Differential) {
    case Forward:
        Super->f_g = DifferentialMethod->ForwardMethod();
        break;
    case Backward:
        Super->f_g = DifferentialMethod->BackwardMethod();
        break;
    case Central:
        Super->f_g = DifferentialMethod->CentralMethod();
        break;
    default:
        break;
    }
    delete DifferentialMethod;
}

void CBasicOptimizer::CSolverMethod::LineSearchMethod() {
    CLineSearch *LineSearch = new CLineSearch(Super);
    LineSearch->Find_Interval();
    switch(Super->Line) {
    case GoldenSection:
        Super->pfX = LineSearch->GoldenSelection();
        break;
    case Fibonacci:
        Super->pfX = LineSearch->Fibonacci();
        break;
    default:
        break;
    }
    delete LineSearch;
}

// calculate normal of gradiant
double CBasicOptimizer::CSolverMethod::norm() {
    double tmp = 0;
    for(int i = 0; i<Super->nDim; i++)
        tmp += Super->f_g[i]*Super->f_g[i];
    return sqrt(tmp);
}

double CBasicOptimizer::CSolverMethod::HestenesStiefelFormula() {
    double tmp1 = 0;
    double tmp2 = 0;
    for(int i = 0; i<Super->nDim; i++) {
        tmp1 += Super->f_g[i] * (Super->f_g[i] - f_g_past[i]);
        tmp2 += f_g_past[i] * (Super->f_g[i] - f_g_past[i]);
    }
    return tmp1/tmp2;
}

double CBasicOptimizer::CSolverMethod::PolakRibiereFormula() {
    double tmp1 = 0;
    double tmp2 = 0;
    for(int i = 0; i<Super->nDim; i++) {
        tmp1 += Super->f_g[i] * (Super->f_g[i] - f_g_past[i]);
        tmp2 += f_g_past[i] * f_g_past[i];
    }
    return tmp1/tmp2;
}

double CBasicOptimizer::CSolverMethod::FletcherReevesFormula() {
    double tmp1 = 0;
    double tmp2 = 0;
    for(int i = 0; i<Super->nDim; i++) {
        tmp1 += Super->f_g[i] * Super->f_g[i];
        tmp2 += f_g_past[i] * f_g_past[i];
    }
    return tmp1/tmp2;
}

double CBasicOptimizer::CSolverMethod::PowellFormula() {
    double tmp1 = 0;
    double tmp2 = 0;
    for(int i = 0; i<Super->nDim; i++) {
        tmp1 += Super->f_g[i]*(Super->f_g[i]-f_g_past[i]);
        tmp2 += f_g_past[i] * f_g_past[i];
    }
    return (tmp1/tmp2)<0 ? 0 : (tmp1/tmp2);
}