Revision 0:e647f6de3d26, committed 2020-03-06

Comitter:

Joeatsumi

Date:

Fri Mar 06 15:03:57 2020 +0000

Commit message:

An software for autopilot(ver2)

Changed in this revision

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/MPU6050.lib	Fri Mar 06 15:03:57 2020 +0000
@@ -0,0 +1,1 @@
+https://os.mbed.com/users/Joeatsumi/code/MPU6050_alter/#44c458576810

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/Matrix/Matrix.cpp	Fri Mar 06 15:03:57 2020 +0000
@@ -0,0 +1,388 @@
+#include "myConstants.h"
+#include "Matrix.h"
+
+
+
+Matrix::Matrix(int row, int col) : row(row), col(col), components(0) {
+    components = new float[row*col];
+    if (!components) error("Memory Allocation Error");
+    for(int i=0; i<row*col; i++) components[i] = 0.0f;
+    if (row == col) {
+        for (int i = 0; i < row; i++) {
+            components[i * col + i] = 1.0f;
+        }
+    }
+}
+
+Matrix::Matrix(int row, int col, float* comps) : row(row), col(col), components(0) {
+    components = new float[row*col];
+    if (!components) error("Memory Allocation Error");
+    memcpy(components, comps, sizeof(float)*row*col);
+}
+
+
+Matrix::~Matrix() {
+    delete[] components;
+}
+
+Matrix::Matrix(const Matrix& m) : row(m.row), col(m.col), components(0) {
+    components = new float[row*col];
+    if (!components) error("Memory Allocation Error");
+    memcpy(components, m.GetpComponents(), sizeof(float)*row*col);
+}
+
+Matrix Matrix::operator-() const{
+    Matrix retMat(*this);
+
+    for (int i = 0; i < row * col; i++) {
+        retMat.components[i] = - this->components[i];
+    }
+
+    return retMat;
+}
+
+Matrix& Matrix::operator=(const Matrix& m) {
+    if (this == &m) return *this;
+    row = m.row;
+    col = m.col;
+    delete[] components;
+    components = new float[row*col];
+    if (!components) error("Memory Allocation Error");
+    memcpy(components, m.GetpComponents(), sizeof(float)*row*col);
+
+    return *this;
+}
+
+Matrix& Matrix::operator+=(const Matrix& m) {
+    if (row != m.GetRow() || col != m.GetCol()) error("Irregular Dimention");
+    
+    for (int i = 0; i < row; i++) {
+        for (int j = 0; j < col; j++) {
+            components[i * col + j] += m.components[i * col + j];
+        }
+    }
+
+    this->CleanUp();
+
+    return *this;
+}
+
+Matrix& Matrix::operator-=(const Matrix& m) {
+    if (row != m.GetRow() || col != m.GetCol()) error("Irregular Dimention");
+
+    for (int i = 0; i < row; i++) {
+        for (int j = 0; j < col; j++) {
+            components[i * col + j] -= m.components[i * col + j];
+        }
+    }
+
+    this->CleanUp();
+
+    return *this;
+}
+/*
+Matrix& Matrix::operator*=(const Matrix& m) {
+    if (col != m.GetRow()) error("Irregular Dimention");
+    Matrix temp = Matrix(*this);
+    
+    col = m.GetCol();
+    delete[] components;
+    components = new float[row*col];
+
+    for (int i = 0; i < row; i++) {
+        for (int j = 0; j < col; j++) {
+            components[i*col + j] = 0.0f;
+            for (int k = 0; k < m.GetRow(); k++) {
+                components[i * col + j] += temp.components[i * col + k] * m.components[k * col + j];
+            }
+        }
+    }
+
+    this->CleanUp();
+
+    return *this;
+}
+*/
+
+Matrix& Matrix::operator*=(float c) {
+    for (int i = 0; i < row; i++) {
+        for (int j = 0; j < col; j++) {
+            components[i*col + j] *= c;
+        }
+    }
+
+    return *this;
+}
+
+Matrix& Matrix::operator/=(float c) {
+    if (fabs(c) < NEARLY_ZERO) error("Division by Zero");
+    for (int i = 0; i < row; i++) {
+        for (int j = 0; j < col; j++) {
+            components[i*col + j] /= c;
+        }
+    }
+
+    return *this;
+}
+
+void Matrix::SetComp(int rowNo, int colNo, float val) {
+    if (rowNo > row || colNo > col) error("Index Out of Bounds Error");
+    components[(rowNo-1)*col + (colNo-1)] = val;
+}
+
+void Matrix::SetComps(float* pComps) {
+    memcpy(components, pComps, sizeof(float) * row * col);
+}
+
+float Matrix::Determinant() const{
+    if (row != col) error("failed to calculate det. : matrix is not square");
+    int decSign = 0;
+    float retVal = 1.0f;
+
+    // 行列のLU分解
+    Matrix LU(this->LU_Decompose(&decSign));
+
+    for (int i = 0; i < LU.row; i++) {
+        retVal *= LU.components[i * LU.col + i];
+    }
+
+    return retVal*decSign;
+}
+
+float Matrix::det() const {
+    if (row != col) error("failed to calculate det : matrix is not square");
+    
+    Matrix temp(*this);
+    int decSign = 1;
+
+    for (int j = 0; j < col - 1; j++) {
+
+        // 列内のみで最大の要素を探す
+        int maxNo = j;
+        for (int k = j; k < row; k++) {
+            if (temp.components[maxNo * col + j] < temp.components[k * col + j]) maxNo = k;
+        }
+        if (maxNo != j) {
+            temp.SwapRow(j + 1, maxNo + 1);
+            decSign *= -1;
+        }
+        // 列内の最大要素が小さ過ぎる場合、行内の最大要素も探す
+        if (fabs(temp.components[j * col + j]) < NEARLY_ZERO) {
+            maxNo = j;
+            for (int k = j; k < col; k++) {
+                if (temp.components[j * col + maxNo] < temp.components[j * col + k])maxNo = k;
+            }
+            if (maxNo != j) {
+                temp.SwapCol(j + 1, maxNo + 1);
+                decSign *= -1;
+            }
+
+            // 列内、行内の最大要素を選んでも小さすぎる場合はエラー
+            if (fabs(temp.components[j * col + j]) < NEARLY_ZERO) {
+                if (row != col) error("failed to calculate det : Division by Zero");
+            }
+        }
+
+        float c1 = 1.0f / temp.components[j * col + j];
+
+        for (int i = j + 1; i < row; i++) {
+            float c2 = temp.components[i * col + j] * c1;
+            for (int k = j; k < col; k++) {
+                temp.components[i * col + k] = temp.components[i * col + k] - c2 * temp.components[j * col + k];
+            }
+        }
+        
+    }
+
+    if (fabs(temp.components[(row - 1) * col + (col - 1)]) < NEARLY_ZERO) return 0.0f;
+
+    float retVal = 1.0f;
+    for (int i = 0; i < row; i++) {
+        retVal *= temp.components[i * col + i];
+    }
+
+    return retVal * decSign;
+}
+
+Matrix Matrix::LU_Decompose(int* sign, Matrix* p) const{
+    if (row != col) error("failed to LU decomposition: matrix is not square");
+    if (sign != 0) *sign = 1;
+    if (p != 0) {
+        if (p->row != row || p->row != p->col) error("failed to LU decomposition: permitation matrix is incorrect");
+        // 置換行列は最初に単位行列にしておく
+        memset(p->components, 0, sizeof(float) * row * col);
+        for (int i = 0; i < row; i++) {
+            p->components[i * col + i] = 1.0f;
+        }
+    }
+    Matrix retVal(*this);
+
+    for (int d = 0; d < row - 1; d++) { // 1行1列ずつ分解を行う
+        // d列目の最大の要素を探索し、見つけた要素の行とd行目を交換する
+        int maxNo = d;
+        for (int i = d; i < row; i++) {
+            if (retVal.components[i * col + d] > retVal.components[maxNo * col + d]) maxNo = i;
+        }
+        if (maxNo != d) {
+            retVal.SwapRow(d + 1, maxNo + 1);
+            if (sign != 0) *sign *= -1;
+            if (p != 0) {
+                p->SwapRow(d + 1, maxNo + 1);
+            }
+        }
+        float c = retVal.components[d * col + d];
+        if (fabs(c) < NEARLY_ZERO) error("failed to LU decomposition: Division by Zero");
+
+        // d行d列目以降の行列について計算
+        for (int i = d+1; i < row; i++) {
+            retVal.components[i * col + d] /= c;
+            for (int j = d+1; j < col; j++) {
+                retVal.components[i * col + j] -= retVal.components[d * col + j] * retVal.components[i * col + d];
+            }
+        }
+    }
+
+    retVal.CleanUp();
+
+    return retVal;
+}
+
+float Matrix::Inverse(Matrix& invm) const{
+    if (row != col) error("failed to get Inv. : matrix is not square");
+
+    Matrix P(*this);
+    Matrix LU(LU_Decompose(0, &P));
+
+    // 分解した行列の対角成分の積から行列式を求める
+    // det = 0　ならfalse
+    float det = 1.0f;
+    for (int i = 0; i < row; i++) {
+        det *= LU.components[i * col + i];
+    }
+    if (fabs(det) < NEARLY_ZERO) {
+        return fabs(det);
+    }
+
+    // U、Lそれぞれの逆行列を計算する
+    Matrix U_inv = Matrix(row, col);
+    Matrix L_inv = Matrix(row, col);
+
+    for (int j = 0; j < col; j++) {
+        for (int i = 0; i <= j; i++) {
+            int i_U = j - i;        // U行列の逆行列は対角成分から上へ向かって
+                                    // 左から順番に値を計算する
+
+            int j_L = col - 1 - j;  // L行列の逆行列は右から順番に
+            int i_L = j_L + i;      // 対角成分から下へ向かって計算する
+
+            if (i_U != j) { // 非対角成分
+                float temp_U = 0.0f;
+                float temp_L = 0.0f;
+
+                for (int k = 0; k < i; k++) {
+
+                    temp_U -= U_inv.components[(j - k) * col + j] * LU.components[i_U * col + (j - k)];
+                    
+                    if (k == 0) {
+                        temp_L -= LU.components[i_L * col + j_L];
+                    } else {
+                        temp_L -= L_inv.components[(j_L + k) * col + j_L] * LU.components[i_L * col + j_L + k];
+                    }
+                    
+                }
+
+                U_inv.components[i_U * col + j] = temp_U / LU.components[i_U * col + i_U];
+                L_inv.components[i_L * col + j_L] = temp_L;
+
+            } else {    // 対角成分
+                if (fabs(LU.components[i_U * col + i_U]) >= NEARLY_ZERO) {
+                    U_inv.components[i_U * col + i_U] = 1.0f / LU.components[i_U * col + i_U];
+                }
+            }
+        }
+    }
+
+    invm = U_inv * L_inv * P;
+
+    return -1.0f;
+}
+
+Matrix Matrix::Transpose() const{
+    //if (row != col) error("failed to get Trans. : matrix is not square");
+    Matrix retVal(col, row);
+
+    for (int i = 0; i < row; i++) {
+        for (int j = 0; j < col; j++) {
+            retVal.components[j * row + i] = this->components[i * col + j];        
+        }
+    }
+
+    return retVal;
+}
+
+Matrix operator+(const Matrix& lhm, const Matrix& rhm) {
+    Matrix temp = Matrix(lhm);
+    temp += rhm;
+    return temp;
+}
+
+Matrix operator-(const Matrix& lhm, const Matrix& rhm) {
+    Matrix temp = Matrix(lhm);
+    temp -= rhm;
+    return temp;
+}
+
+Matrix operator*(const Matrix& lhm, const Matrix& rhm) {
+    if(lhm.GetCol() != rhm.GetRow()) error("Matrix product Error: Irregular Dimention.");
+    int row = lhm.GetRow();
+    int col = rhm.GetCol();
+    int sum = lhm.GetCol();
+    Matrix temp(row, col);
+    
+    for (int i = 1; i <= row; i++) {
+        for (int j = 1; j <= col; j++) {
+            float temp_c = 0.0f;
+            for (int k = 1; k <= sum; k++) {
+                temp_c += lhm.GetComp(i, k) * rhm.GetComp(k, j);
+            }
+            temp.SetComp(i, j, temp_c);
+        }
+    }
+    
+    return temp;
+}
+
+void Matrix::CleanUp() {
+    int num = row*col;
+    float maxComp = 0.0f;
+    for (int i = 0; i < num; i++) {
+        if (maxComp < fabs(components[i])) maxComp = fabs(components[i]);
+    }
+    if (maxComp > NEARLY_ZERO) {
+        for (int i = 0; i < num; i++) {
+            if (fabs(components[i]) / maxComp < ZERO_TOLERANCE) components[i] = 0.0f;
+        }
+    }
+}
+
+void Matrix::SwapRow(int rowNo1, int rowNo2) {
+    if (rowNo1 > row || rowNo2 > row) error("Index Out of Bounds Error !!");
+    float* temp = new float[col];
+
+    memcpy(temp, components + (rowNo1 - 1) * col, sizeof(float) * col);
+    memcpy(components + (rowNo1 - 1) * col, components + (rowNo2 - 1) * col, sizeof(float) * col);
+    memcpy(components + (rowNo2 - 1) * col, temp, sizeof(float) * col);
+
+    delete[] temp;
+}
+
+void Matrix::SwapCol(int colNo1, int colNo2) {
+    if (colNo1 > col || colNo2 > col) error("Index Out of Bounds Error !!");
+    float temp = 0.0f;
+
+    for (int i = 0; i < row; i++) {
+        temp = components[i * col + colNo1];
+        components[i * col + colNo1] = components[i * col + colNo2];
+        components[i * col + colNo2] = temp;
+    }
+}
\ No newline at end of file

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/Matrix/Matrix.h	Fri Mar 06 15:03:57 2020 +0000
@@ -0,0 +1,115 @@
+#pragma once
+#include "mbed.h"
+
+class Matrix
+{
+public:
+    /********** コンストラクタ　デストラクタ **********/
+    Matrix(int row, int col);
+    Matrix(int row, int col, float* comps);
+    ~Matrix();
+    Matrix(const Matrix& m);
+
+    /********** メンバ演算子 **********/
+    Matrix operator-() const;
+    Matrix& operator=(const Matrix& m);
+    Matrix& operator+=(const Matrix& m);
+    Matrix& operator-=(const Matrix& m);
+    //Matrix& operator*=(const Matrix& m);
+    Matrix& operator*=(float c);
+    Matrix& operator/=(float c);
+
+    /********** その他関数 **********/
+    /*
+        行列の成分を設定
+        引数：rowNo 行番号
+              colNo 列番号
+              val 設定値
+    */
+    void SetComp(int rowNo, int colNo, float val);
+
+    /*
+        行列の成分を全て設定。全成分を一度に指定する必要がある。
+        引数：pComps 設定値の入ったfloat配列。
+    */
+    void SetComps(float* pComps);
+    /*
+        行列式を計算する。行列が正方行列で無い場合にはエラー。
+    */
+    float Determinant() const;
+    
+    /*
+        行列式を計算する。行列が正方行列で無い場合にはエラー。
+    */
+    float det() const;
+
+    /*
+        行列をLU分解する
+        引数：sign (省略可)置換操作の符号を格納するポインタ
+             p (省略可)置換行列を格納する行列のポインタ。（分解する行列と同じ列数の正方行列)
+        返り値：LU分解後の行列。下三角要素がL、対角・上三角要素がUに対応する。
+    */
+    Matrix LU_Decompose(int* sign = 0, Matrix * p = 0) const;
+    
+    /*
+        逆行列を生成する
+        返り値で逆行列が存在するか否かを判断
+        引数：逆行列を格納する行列
+        返り値：逆行列が存在するか否か
+    */
+    float Inverse(Matrix& invm) const;
+    
+    /*
+        転置行列を生成する
+        返り値：転置行列
+    */
+    Matrix Transpose() const;
+    
+    /*
+        行列の行の入れ替えを行う
+        引数：rowNo1　行番号１
+             rowNo2　行番号２
+    */
+    void SwapRow(int rowNo1, int rowNo2);
+    /*
+        行列の列の入れ替えを行う
+        引数：colNo1　列番号１
+             colNo2　列番号２
+    */
+    void SwapCol(int colNo1, int colNo2);
+    
+    /********** インライン関数 **********/
+    inline int GetRow() const {
+        return row;
+    }
+
+    inline int GetCol() const {
+        return col;
+    }
+
+    inline const float* GetpComponents() const {
+        return (const float*)components;
+    }
+
+    inline float GetComp(int rowNo, int colNo) const {
+        if (rowNo > row || colNo > col) error("Index Out of Bounds Error !!");
+        return components[(rowNo-1)*col + (colNo-1)];
+    }
+
+private:
+    int row;
+    int col;
+    float* components;
+
+    /*
+        行列の成分の中で無視できるほど小さい値を0と置き換える（掃除する）
+    */
+    void CleanUp();
+
+};
+
+// グローバル演算子
+Matrix operator+(const Matrix& lhm, const Matrix& rhm);
+Matrix operator-(const Matrix& lhm, const Matrix& rhm);
+Matrix operator*(const Matrix& lhm, const Matrix& rhm);
+

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/SDFileSystem.lib	Fri Mar 06 15:03:57 2020 +0000
@@ -0,0 +1,1 @@
+http://mbed.org/users/mbed_official/code/SDFileSystem/#8db0d3b02cec

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/Vector/Vector.cpp	Fri Mar 06 15:03:57 2020 +0000
@@ -0,0 +1,177 @@
+#include "myConstants.h"
+#include "Vector.h"
+
+
+Vector::Vector(int dim) : dim(dim), components(0){
+    components = new float[dim];
+    if (!components) error("Memory Allocation Error");
+    for(int i=0; i<dim; i++) components[i] = 0.0f;
+}
+
+
+Vector::~Vector() {
+    delete[] components;
+}
+
+Vector::Vector(const Vector& v) : dim(v.dim), components(0) {
+    components = new float[dim];
+    if (!components) error("Memory Allocation Error");
+    memcpy(components, v.GetpComponents(), sizeof(float)*dim);
+}
+
+Vector& Vector::operator=(const Vector& v) {
+    if (this == &v) return *this;
+    dim = v.dim;
+    delete[] components;
+    components = new float[dim];
+    if (!components) error("Memory Allocation Error");
+    memcpy(components, v.GetpComponents(), sizeof(float)*dim);
+
+    return *this;
+}
+
+Vector Vector::operator+() {
+    return *this;
+}
+
+Vector Vector::operator-() {
+    Vector retVec(*this);
+    retVec *= -1;
+    return retVec;
+}
+
+Vector& Vector::operator*=(float c) {
+    for (int i = 0; i < dim; i++) {
+        components[i] *= c;
+    }
+
+    return *this;
+}
+
+Vector& Vector::operator/=(float c) {
+    if (fabs(c) < NEARLY_ZERO) error("Division by Zero");
+    for (int i = 0; i < dim; i++) {
+        components[i] /= c;
+    }
+
+    return *this;
+}
+
+Vector& Vector::operator+=(const Vector& v) {
+    if (dim != v.dim) error("failed to add: Irregular Dimention");
+    for (int i = 0; i < dim; i++) {
+        components[i] += v.components[i];
+    }
+
+    this->CleanUp();
+
+    return *this;
+}
+
+Vector& Vector::operator-=(const Vector& v) {
+    if (dim != v.dim) error("failed to subtract: Irregular Dimention");
+    for (int i = 0; i < dim; i++) {
+        components[i] -= v.components[i];
+    }
+
+    this->CleanUp();
+
+    return *this;
+}
+
+void Vector::SetComp(int dimNo, float val) {
+    if (dimNo > dim) error("Index Out of Bounds Error");
+    components[dimNo-1] = val;
+}
+
+void Vector::SetComps(float* pComps) {
+    memcpy(components, pComps, sizeof(float) * dim);
+}
+
+float Vector::GetNorm() const {
+    float norm = 0.0f;
+    for (int i = 0; i < dim; i++) {
+        norm += components[i] * components[i];
+    }
+    return sqrt(norm);
+}
+
+Vector Vector::Normalize() const {
+    float norm = GetNorm();
+    Vector temp(*this);
+    for (int i = 0; i < dim; i++) {
+        temp.components[i] /= norm;
+    }
+    temp.CleanUp();
+    return temp;
+}
+
+Vector Vector::GetParaCompTo(Vector v) {
+    Vector norm_v = v.Normalize();
+    return (*this * norm_v) * norm_v;
+}
+
+Vector Vector::GetPerpCompTo(Vector v) {
+    return (*this - this->GetParaCompTo(v));
+}
+
+void Vector::CleanUp() {
+    float maxComp = 0.0f;
+    for (int i = 0; i < dim; i++) {
+        if (fabs(components[i]) > maxComp) maxComp = fabs(components[i]);
+    }
+    if (maxComp > NEARLY_ZERO) {
+        for (int i = 0; i < dim; i++) {
+            if (fabs(components[i]) / maxComp < ZERO_TOLERANCE) components[i] = 0.0f;
+        }
+    }
+}
+
+Vector operator+(const Vector& lhv, const Vector& rhv) {
+    Vector retVec(lhv);
+    retVec += rhv;
+    return retVec;
+}
+
+Vector operator-(const Vector& lhv, const Vector& rhv) {
+    Vector retVec(lhv);
+    retVec -= rhv;
+    return retVec;
+}
+
+Vector Cross(const Vector& lhv, const Vector& rhv) {
+    if (lhv.GetDim() != 3) error("failed to cross: variable 'dim' must be 3");
+    if (lhv.GetDim() != rhv.GetDim()) error("failed to cross: Irregular Dimention");
+
+    Vector retVec(lhv.GetDim());
+
+    for (int i = 0; i < lhv.GetDim(); i++) {
+        retVec.SetComp(i + 1, lhv.GetComp((i + 1) % 3 + 1) * rhv.GetComp((i + 2) % 3 + 1)
+            - lhv.GetComp((i + 2) % 3 + 1) * rhv.GetComp((i + 1) % 3 + 1));
+    }
+
+    return retVec;
+}
+
+Vector operator*(const float c, const Vector& rhv) {
+    Vector retVec(rhv);
+    retVec *= c;
+    return retVec;
+}
+
+Vector operator*(const Vector& lhv, const float c) {
+    Vector retVec(lhv);
+    retVec *= c;
+    return retVec;
+}
+
+float operator*(const Vector& lhv, const Vector& rhv) {
+    if (lhv.GetDim() != rhv.GetDim()) error("Irregular Dimention");
+    float retVal = 0.0f;
+
+    for (int i = 1; i <= lhv.GetDim(); i++) {
+        retVal += lhv.GetComp(i) * rhv.GetComp(i);
+    }
+
+    return retVal;
+}

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/Vector/Vector.h	Fri Mar 06 15:03:57 2020 +0000
@@ -0,0 +1,56 @@
+#pragma once
+#include "mbed.h"
+
+class Matrix;
+
+class Vector {
+public:
+    Vector(int dim);
+    ~Vector();
+    Vector(const Vector& v);
+
+    Vector& operator=(const Vector& v);
+    Vector operator+();
+    Vector operator-();
+    Vector& operator*=(float c);
+    Vector& operator/=(float c);
+    Vector& operator+=(const Vector& v);
+    Vector& operator-=(const Vector& v);
+
+    void SetComp(int dimNo, float val);
+    void SetComps(float* vals);
+    float GetNorm() const;
+    Vector Normalize() const;
+    Vector GetParaCompTo(Vector v);
+    Vector GetPerpCompTo(Vector v);
+
+    inline int GetDim() const {
+        return dim;
+    }
+
+    inline const float* GetpComponents() const {
+        return (const float*)components;
+    }
+
+    inline float GetComp(int dimNo) const {
+        if (dimNo > dim) error("Index Out of Bounds Error !!");
+        return components[dimNo-1];
+    }
+
+    void CleanUp();
+
+private:
+    int dim;
+    float* components;
+
+    Vector& operator*=(const Matrix& m);
+    Vector& operator*=(const Vector& m);
+};
+
+Vector operator+(const Vector& lhv, const Vector& rhv);
+Vector operator-(const Vector& lhv, const Vector& rhv);
+Vector Cross(const Vector& lhv, const Vector& rhv);
+Vector operator*(const float c, const Vector& rhv);
+Vector operator*(const Vector& lhv, const float c);
+float operator*(const Vector& lhv, const Vector& rhv);
+

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/Vector/Vector_Matrix_operator.cpp	Fri Mar 06 15:03:57 2020 +0000
@@ -0,0 +1,35 @@
+#include "Vector_Matrix_operator.h"
+
+Vector operator*(const Matrix& lhm, const Vector& rhv) {
+    if (lhm.GetCol() != rhv.GetDim()) error("Irregular Dimention");
+    Vector retVec(lhm.GetRow());
+
+    for (int i = 1; i <= lhm.GetRow(); i++) {
+        float temp = 0.0f;
+        for (int j = 1; j <= rhv.GetDim(); j++) {
+            temp += lhm.GetComp(i, j)*rhv.GetComp(j);
+        }
+        retVec.SetComp(i, temp);
+    }
+
+    retVec.CleanUp();
+
+    return retVec;
+}
+
+Vector operator*(const Vector& lhv, const Matrix& rhm) {
+    if (lhv.GetDim() != rhm.GetRow()) error("Irregular Dimention");
+    Vector retVec(rhm.GetCol());
+
+    for (int i = 1; i <= rhm.GetCol(); i++) {
+        float temp = 0.0f;
+        for (int j = 1; j <= lhv.GetDim(); j++) {
+            temp += lhv.GetComp(j) * rhm.GetComp(j, i);
+        }
+        retVec.SetComp(i, temp);
+    }
+
+    retVec.CleanUp();
+
+    return retVec;
+}
\ No newline at end of file

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/Vector/Vector_Matrix_operator.h	Fri Mar 06 15:03:57 2020 +0000
@@ -0,0 +1,5 @@
+#include "Matrix.h"
+#include "Vector.h"
+
+Vector operator*(const Matrix& lhm, const Vector& rhv);
+Vector operator*(const Vector& lhv, const Matrix& rhm);
\ No newline at end of file

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/main.cpp	Fri Mar 06 15:03:57 2020 +0000
@@ -0,0 +1,505 @@
+//==================================================
+//Auto pilot(prototype2)
+//
+//MPU board: mbed LPC1768
+//Multiplexer TC74HC157AP
+//Accelerometer +Gyro sensor : GY-521
+//2019/11/17 A.Toda
+//==================================================
+#include "mbed.h"
+#include "MPU6050.h"
+#include "SDFileSystem.h"
+
+#include "Vector.h"
+#include "Matrix.h"
+#include "Vector_Matrix_operator.h"
+
+#include "math.h"
+
+//==================================================
+#define RAD_TO_DEG          57.2957795f             // 180 / π
+#define MAX_MEAN_COUNTER 500
+
+#define ACC_X -1.205178682//offset of x-axi accelerometer
+#define ACC_Y -0.141728488//offset of y-axi accelerometer
+#define ACC_Z -0.339272785//offset of z-axi accelerometer
+
+#define ACC_GAIN_X 1.0
+#define ACC_GAIN_Y 0.995906146
+#define ACC_GAIN_Z 1.017766039
+
+#define THRESHOLD_PWM 0.0015
+#define M_A_PWM 0.0017
+#define SERVO_PERIOD  0.020
+
+#define PITCH_TARGET 0.0
+
+/*
+8/2/2020でのゲイン
+#define P_P_GAIN 3.0
+#define P_I_GAIN 0.5
+#define P_D_GAIN 1.0
+*/
+
+#define OFFSET_PWM_ELE 0.0015
+#define MAX_PWM_ELE 0.00175
+#define MIN_PWM_ELE 0.00125
+
+#define LPF_ELE_K 0.110//この値が小さいとローパスフィルタが強くなる 0.050は小さすぎる
+                        //0.200は動きが機敏過ぎる。
+//==================================================
+
+//Port Setting
+SDFileSystem sd(p5, p6, p7, p8, "sd");//pins for sd slot 
+MPU6050 mpu(p9, p10);  //Accelerometer + Gyro
+                        //(SDA,SCLK)
+DigitalIn logging_terminater(p16);
+InterruptIn reading_port(p18); 
+DigitalOut mux_switch(p19);
+PwmOut ELE(p21);
+
+Serial pc(USBTX, USBRX);    //UART
+
+//Pointer of sd card
+FILE *fp;
+
+//==================================================
+//Accelerometer and gyro data
+//==================================================
+double acc[3]; //variables for accelerometer
+double gyro[3]; //variables for gyro
+
+double offset_gyro_x=0.0;
+double offset_gyro_y=0.0;
+
+double sum_gyro_x=0.0;
+double sum_gyro_y=0.0;
+
+double threshold_acc,threshold_acc_ini;
+
+double dev  =0.0;
+double i_dev=0.0;
+double d_dev=0.0;
+
+double old_i_dev=0.0;
+//==================================================
+//Atitude data
+//==================================================
+double roll_and_pitch_acc[2];//atitude from acceleromter
+double roll_and_pitch[2];//atitude from gyro and acceleromter
+
+/*--------------------------行列、ベクトル-----------------------------*/
+Matrix rate_angle_matrix(4,4);//角速度行列 クォータニオンの更新に使う
+Vector quaternion(4),pre_quaternion(4),dump_1(4);;//クォータニオン
+/*-------------------------------------------------------------------*/
+
+//==================================================
+//Timer valiables
+//==================================================
+Timer ch_time;//timer for calculate pulse width
+Timer passed_time;//timer for calculate atitude
+
+double measured_pre_pulse=0.0;
+double measured_pulse=0.0;
+
+double time_new;
+double time_old;
+
+double pulse_width_ele,deflection_ele,old_deflection_ele;
+
+//=================================================
+//PID Gain
+//==================================================
+float P_P_GAIN,P_I_GAIN,P_D_GAIN;
+int P_GAIN,I_GAIN,D_GAIN;
+
+//=================================================
+//エレベータの舵角
+//=================================================
+double pitch_command;
+
+//=================================================
+//手動か自動かのインジケータ0なら手動、1なら自動
+//=================================================
+int m_a_indicater;
+//=================================================
+//Functions for rising and falind edge interrution
+//=================================================
+//rise edge
+void rising_edge(){
+    ch_time.reset();//reset timer counter
+    measured_pre_pulse=ch_time.read();
+    
+}
+
+//falling edge
+void falling_edge(){
+    
+    measured_pre_pulse=(ch_time.read()-measured_pre_pulse);
+    //pc.printf("The pulse width=%f\r\n",measured_pre_pulse);
+    if(measured_pre_pulse>M_A_PWM){
+        mux_switch=1;
+        m_a_indicater=1;//set indicater as auto
+        }else{
+            mux_switch=0;
+            m_a_indicater=0;//set indicater as manual
+            }
+}
+
+//terminate logging
+void end_of_log(){
+     //flipper.detach();
+     fclose(fp);//close "Atitude_angles.csv"
+     pc.printf("Logging was terminated.");
+     
+    }
+//==================================================
+//Gyro and accelerometer functions
+//==================================================
+//get data
+void  aquisition_sensor_values(double *a,double *g){
+    
+    float ac[3],gy[3];
+    
+    mpu.getAccelero(ac);//get acceleration (Accelerometer)
+                                //x_axis acc[0]
+                                //y_axis acc[1]
+                                //z_axis acc[2]
+    mpu.getGyro(gy);   //get rate of angle(Gyro)
+                      //x_axis gyro[0]
+                      //y_axis gyro[1]
+                      //z_axis gyro[2]
+                      
+    //Invertion for direction of Accelerometer axis
+    ac[0]*=(-1.0);
+    ac[2]*=(-1.0);
+    
+    ac[0]=(ac[0]-ACC_X)/ACC_GAIN_X;
+    ac[1]=(ac[1]-ACC_Y)/ACC_GAIN_Y;
+    ac[2]=(ac[2]-ACC_Z)/ACC_GAIN_Z;
+            
+    //Unit convertion of rate of angle(radian to degree)
+    gy[0]*=RAD_TO_DEG;
+    gy[0]*=(-1.0);
+        
+    gy[1]*=RAD_TO_DEG;        
+    gy[2]*=RAD_TO_DEG;
+    gy[2]*=(-1.0);
+  
+    for(int i=0;i<3;i++){
+        a[i]=double(ac[i]);
+        g[i]=double(gy[i]);
+        }
+    g[0]-=offset_gyro_x;//offset rejection
+    g[1]-=offset_gyro_y;//offset rejection
+    
+    return;
+    
+}
+
+//calculate offset of gyro
+void offset_calculation_for_gyro(){
+    
+    //Accelerometer and gyro setting 
+    mpu.setAcceleroRange(0);//acceleration range is +-2G
+    mpu.setGyroRange(1);//gyro rate is +-500degree per second(dps)
+    
+    //calculate offset of gyro
+    for(int mean_counter=0; mean_counter<MAX_MEAN_COUNTER ;mean_counter++){
+        aquisition_sensor_values(acc,gyro);
+        sum_gyro_x+=gyro[0];
+        sum_gyro_y+=gyro[1];
+        wait(0.01);
+        }
+    
+    offset_gyro_x=sum_gyro_x/MAX_MEAN_COUNTER;
+    offset_gyro_y=sum_gyro_y/MAX_MEAN_COUNTER;
+    
+    return;
+}
+
+//atitude calculation from acceleromter
+void atitude_estimation_from_accelerometer(double *a,double *roll_and_pitch){
+    
+    roll_and_pitch[0] = atan(a[1]/a[2])*RAD_TO_DEG;//roll
+    roll_and_pitch[1] = atan(a[0]/sqrt( (a[1]*a[1]+a[2]*a[2]) ) )*RAD_TO_DEG;//pitch
+    
+    return;
+}
+
+//quaternion to euler
+void quaternion_to_euler(Vector& qua,double *roll_and_pitch ){
+    
+    double q0=double (qua.GetComp(1));
+    double q1=double (qua.GetComp(2));
+    double q2=double (qua.GetComp(3));
+    double q3=double (qua.GetComp(4));
+    
+    roll_and_pitch[0]=atan((q2*q3+q0*q1)/(q0*q0-q1*q1-q2*q2+q3*q3)) ;//roll
+    roll_and_pitch[1]=-asin(2*(q1*q3-q0*q2));
+    
+    return;
+    }
+
+//quaternion to euler
+void euler_to_quaternion(Vector& qua,double *roll_and_pitch ){
+    
+    double roll_rad_2=(roll_and_pitch[0]/57.3)/2.0;
+    double pitch_rad_2=(roll_and_pitch[1]/57.3)/2.0;
+    double yaw_rad_2=0.0;
+    
+    float q0= cos(roll_rad_2)*cos(pitch_rad_2)*cos(yaw_rad_2)
+                +sin(roll_rad_2)*sin(pitch_rad_2)*sin(yaw_rad_2);
+                
+    float q1= sin(roll_rad_2)*cos(pitch_rad_2)*cos(yaw_rad_2)
+                -cos(roll_rad_2)*sin(pitch_rad_2)*sin(yaw_rad_2);
+                
+    float q2= cos(roll_rad_2)*sin(pitch_rad_2)*cos(yaw_rad_2)
+                +sin(roll_rad_2)*cos(pitch_rad_2)*sin(yaw_rad_2);
+
+    float q3= cos(roll_rad_2)*cos(pitch_rad_2)*sin(yaw_rad_2)
+                -sin(roll_rad_2)*sin(pitch_rad_2)*cos(yaw_rad_2);
+
+    //クォータニオン行列の作成（クォータニオンの演算に用いる）
+    float quaternion_elements[4]={q0,q1,q2,q3}; 
+    qua.SetComps(quaternion_elements);
+    
+    
+    return;
+    }
+
+//atitude calculation
+void atitude_update(){
+    
+    //慣性センサの計測
+    aquisition_sensor_values(acc,gyro);
+    //角速度行列の作成（クォータニオンの演算に用いる）
+    float rate_angles[16]={0.0,-float(gyro[0]),-float(gyro[1]),-float(gyro[2])
+                          ,float(gyro[0]),0.0,float(gyro[2]),-float(gyro[1])
+                          ,float(gyro[1]),-float(gyro[2]),0.0,float(gyro[0])
+                          ,float(gyro[0]),float(gyro[1]),-float(gyro[2]),0.0}; 
+                          
+    rate_angle_matrix.SetComps(rate_angles);
+    //クォータニオンの演算
+    //pc.printf("クォータニオンの演算\r\n");
+    float coefficents=0.5*(time_new-time_old);
+    float coefficents_elements[16]={coefficents,0.0,0.0,0.0
+                                      ,0.0,coefficents,0.0,0.0
+                                      ,0.0,0.0,coefficents,0.0
+                                      ,0.0,0.0,0.0,coefficents};
+    Vector coefficents_vector(4);
+    coefficents_vector.SetComps(coefficents_elements);
+    
+    dump_1=rate_angle_matrix*coefficents_vector;
+    quaternion=dump_1+pre_quaternion;
+    
+    //正規化
+    //pc.printf("正規化\r\n");
+    quaternion=quaternion.Normalize();
+    //クォータニオンからオイラー角への変換
+    quaternion_to_euler(quaternion,roll_and_pitch_acc );
+    
+    threshold_acc=sqrt(acc[0]*acc[0]+acc[1]*acc[1]+acc[2]*acc[2]);
+    
+    if((threshold_acc>=0.9*threshold_acc_ini)
+          &&(threshold_acc<=1.1*threshold_acc_ini)){
+        
+        atitude_estimation_from_accelerometer(acc,roll_and_pitch_acc);
+        roll_and_pitch[0] = 0.98*roll_and_pitch[0] + 0.02*roll_and_pitch_acc[0];
+        roll_and_pitch[1] = 0.98*roll_and_pitch[1] + 0.02*roll_and_pitch_acc[1];
+        
+        }else{}
+    //補正したオイラー角をクォータニオンへ変換
+    euler_to_quaternion(pre_quaternion,roll_and_pitch_acc );
+    
+    //microSDに記録する
+    //経過時間,ロール角,ピッチ角,操縦方式,慣性センサの値
+    fprintf(fp, "%f,%f,%f,%d,%f,%f,%f,%f,%f,%f\r\n"
+        ,time_new,roll_and_pitch[0],roll_and_pitch[1],m_a_indicater,gyro[0],gyro[1],acc[0],acc[1],acc[2],pitch_command);
+    
+    return;
+
+}
+
+//elevation commnad to PWM
+double elevation_to_PWM(double elevation_command){
+    
+    /*
+    PWM信号0.25msが舵角の16.45度に相当する.
+    */
+    
+    double PWM_pitch = (elevation_command*1.519)/100000+ OFFSET_PWM_ELE;
+    //double PWM_pitch = ( ((elevation_command)*6.0/1000.0)/1000.0  )+ OFFSET_PWM_ELE;
+    
+    /*PWMコマンドの上限と下限の設定*/
+    if(PWM_pitch>MAX_PWM_ELE){
+        PWM_pitch=MAX_PWM_ELE;
+        
+        }else if(PWM_pitch<MIN_PWM_ELE){
+            PWM_pitch=MIN_PWM_ELE;
+            }
+    
+    return PWM_pitch;
+    
+    }
+
+double deflection_of_ele(double pitch){
+    
+    double add_deflection=((pitch-PITCH_TARGET)*6.0/1000.0)/1000.0;
+    
+    return add_deflection;
+    }
+
+//PID controller
+double pitch_PID_controller(double pitch,double target,double gyro_pitch){
+    
+    dev=target-pitch;
+    
+    //アンチワインドアップ
+    i_dev=old_i_dev+dev*(time_new-time_old);
+    if(i_dev>=25.0){
+        i_dev=25.0;
+    }else if(i_dev<=-25.0){
+        i_dev=-25.0;
+    }
+    //アンチワインドアップ終わり
+            
+    old_i_dev=i_dev;
+    
+    d_dev=-gyro_pitch;
+    
+    pitch_command = double(P_P_GAIN*dev+P_I_GAIN*i_dev+P_D_GAIN*d_dev);
+    
+    double pwm_command = elevation_to_PWM(pitch_command);//pwm信号に変換
+    
+    return pwm_command;
+
+    }
+
+//LPF
+
+double LPF_pitch(double c_com,double old_com){
+    
+    double lpf_output=(1-LPF_ELE_K)*old_com+LPF_ELE_K*c_com;
+    
+    return lpf_output;
+    }
+
+//==================================================
+//Main
+//==================================================
+int main() {
+
+    wait(5.0);
+    
+    //UART initialization
+    pc.baud(115200);
+    
+    //define servo period
+    ELE.period(SERVO_PERIOD);  // servo requires a 20ms period
+    pulse_width_ele=0.0015;
+    
+    //timer starts
+    ch_time.start();
+    passed_time.start();
+    
+    time_old=0.0;
+    
+    //declare interrupitons
+    reading_port.rise(rising_edge);
+    reading_port.fall(falling_edge);
+    
+    /*
+    mux_switch=0;//set circit as manual mode
+    m_a_indicater=0;//set indicater as manual
+    */
+    
+    //gyro and accelerometer initialization
+    offset_calculation_for_gyro();
+    
+    //determine initilal atitude
+    aquisition_sensor_values(acc,gyro);
+    atitude_estimation_from_accelerometer(acc,roll_and_pitch);
+    euler_to_quaternion(pre_quaternion,roll_and_pitch);
+    
+    threshold_acc_ini=sqrt(acc[0]*acc[0]+acc[1]*acc[1]+acc[2]*acc[2]);
+    
+    /*
+    PIDゲインをsdカードのテキストファイルから読み取る
+    int型しか読みとれないので、希望する値の10倍をテキストファイルに書き込んでいる。
+    Pゲインを3.0,Iゲインを0.5,Dゲインを1.0とする場合
+    
+    30,5,10
+    
+    のようにテキストデータをsdカードに用意する。  
+    */
+    //open PID gain file in sd card 
+    FILE*ga = fopen("/sd/Gain_data.txt", "r");
+    if(ga == NULL) {
+        error("Could not open file for write\n");
+    }
+    
+    while (!feof(ga)) {
+        int n = fscanf(ga, "%d,%d,%d",&P_GAIN, &I_GAIN, &D_GAIN);
+        
+        if(n!= 3){
+            error("Could not read 3 elements");
+            }
+    }
+    
+    P_P_GAIN=P_GAIN/10.0;
+    P_I_GAIN=I_GAIN/10.0;
+    P_D_GAIN=D_GAIN/10.0;
+    
+    fclose(ga);
+    
+    //create folder(sd) in sd card
+    mkdir("/sd", 0777);
+    //create "Atitude_angles.csv" in folder(sd)
+    fp = fopen("/sd/Atitude_angles.csv", "a");//ファイルがある場合は追加で書き込み
+        
+    if(fp == NULL) {
+        error("Could not open file for write\n");
+    }
+    
+    //Logging starts
+    pc.printf("Logging starts.");
+    
+    //write PID gain on sd card 
+    fprintf(fp,"%f,%f,%f\r\n",P_P_GAIN,P_I_GAIN,P_D_GAIN);
+    
+    //while
+    while(1){
+         
+         if(logging_terminater==1){
+             end_of_log();
+             }else{}
+         
+         time_new=passed_time.read();
+         
+         atitude_update();
+        
+         time_old=time_new;
+         
+         /*
+         ここから先でサーボの操舵角を姿勢角に応じて変化させる。関数pitch_PID_controller
+         はpitch角とpitch角速度に応じてサーボのパルス幅を返す関数である。
+         */
+         
+         deflection_ele = pitch_PID_controller(roll_and_pitch[1],PITCH_TARGET,gyro[1]);
+         
+         //LPF
+         deflection_ele = LPF_pitch(deflection_ele,old_deflection_ele);
+         old_deflection_ele = deflection_ele;
+         
+         //servo output
+         ELE.pulsewidth(deflection_ele);
+         
+         //PCにつないでデバッグを行う際に表示する
+         pc.printf("%f,%f,%f,%f,%d\r\n",time_new,deflection_ele,roll_and_pitch[0],roll_and_pitch[1],m_a_indicater);
+         
+         wait(0.002);
+        
+    }//while ends
+    
+}//main ends
\ No newline at end of file

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/mbed.bld	Fri Mar 06 15:03:57 2020 +0000
@@ -0,0 +1,1 @@
+https://os.mbed.com/users/mbed_official/code/mbed/builds/e1686b8d5b90
\ No newline at end of file

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/myConstants.h	Fri Mar 06 15:03:57 2020 +0000
@@ -0,0 +1,37 @@
+#pragma once
+
+/* Math Constants */
+#define NEARLY_ZERO         0.000000001f
+#define ZERO_TOLERANCE      0.000001f
+#define RAD_TO_DEG          57.2957795f             // 180 / π
+#define DEG_TO_RAD          0.0174532925f           // π / 180
+
+/* Accelerometer */
+#define ACC_LSB_TO_G        0.0000610351562f        // g/LSB (1/2^14
+#define G_TO_MPSS           9.8f                    // (m/s^2)/g
+
+/* Gyro Sensor */
+//#define GYRO_LSB_TO_DEG     0.0304878048f           // deg/LSB (1/32.8
+#define GYRO_LSB_TO_DEG     0.0152671755f           // deg/LSB (1/65.5
+//#define GYRO_LSB_TO_DEG     0.00763358778f          // deg/LSB (1/131
+
+/* Pressure Sensor */
+#define PRES_LSB_TO_HPA     0.000244140625f         // hPa/LSB (1/4096
+
+inline float TempLsbToDeg(short int temp) {
+    return (42.5f + (float)temp * 0.00208333333f);  // degree_C = 42.5 + temp / 480;
+}
+
+/* GPS */
+#define GPS_SQ_E                0.00669437999f      // （第一離心率）^2
+#define GPS_A                   6378137.0f          // 長半径(赤道半径)(m)
+#define GPS_B                   6356752.3f          // 短半径(極半径)(m)
+
+/* Geomagnetic Sensor */
+#define MAG_LSB_TO_GAUSS    0.00092f                // Gauss/LSB
+#define MAG_MAGNITUDE       0.46f                   // Magnitude of GeoMagnetism (Gauss)
+#define MAG_SIN             -0.754709580f           // Sin-Value of Inclination
+#define MAG_DECLINATION     7.5f                    // declination (deg)
+
+/* ADC */
+#define ADC_LSB_TO_V        0.000050354f            // 3.3(V)/65535(LSB)
\ No newline at end of file