Gerrit Pathuis
Example of a least square caculation, Matrix calculation based on CMSIS-DSP library, x(t)=a+b*t+c*t^2.
Comment out line 226 of "arm_math.h" to prevent a lot of compiler warnings
Revision 0:9087ea0ad689, committed 2015-03-10
- Comitter:
- GerritPathuis
- Date:
- Tue Mar 10 12:14:31 2015 +0000
- Commit message:
- Program works
Changed in this revision
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/dsp.lib Tue Mar 10 12:14:31 2015 +0000 @@ -0,0 +1,1 @@ +http://mbed.org/users/simon/code/dsp/#208cacc9d789
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/main.cpp Tue Mar 10 12:14:31 2015 +0000
@@ -0,0 +1,88 @@
+///////////////////////////////////////
+// FRDM-K22F //
+// Least square calculation (Matrix) //
+// Based on the DSP libtary //
+// Digital Signal Processing //
+// //
+// Results can be read via Tera Term //
+///////////////////////////////////////
+#include "mbed.h"
+#include "arm_math.h"
+
+#define NUMSAMPLES 51
+#define NUMUNKNOWS 3
+
+Serial pc(USBTX, USBRX); // tx, rx
+
+// time is evenly spaced bit this in not required
+float32_t tData[NUMSAMPLES] = {
+ 0.0f, 0.1f, 0.2f, 0.3f, 0.4f, 0.5f, 0.6f, 0.7f,
+ 0.8f, 0.9f, 1.0f, 1.1f, 1.2f, 1.3f, 1.4f, 1.5f,
+ 1.6f, 1.7f, 1.8f, 1.9f, 2.0f, 2.1f, 2.2f, 2.3f,
+ 2.4f, 2.5f, 2.6f, 2.7f, 2.8f, 2.9f, 3.0f, 3.1f,
+ 3.2f, 3.3f, 3.4f, 3.5f, 3.6f, 3.7f, 3.8f, 3.9f,
+ 4.0f, 4.1f, 4.2f, 4.3f, 4.4f, 4.5f, 4.6f, 4.7f,
+ 4.8f, 4.9f, 5.0f
+};
+
+float32_t xData[NUMSAMPLES] = {
+ 07.4213f, 21.7231f, -7.2828f, 21.2254f, 20.2221f, 10.3585f, 20.2033f, 29.2690f,
+ 57.7152f, 53.6075f, 22.8209f, 59.8714f, 43.1712f, 38.4436f, 46.0499f, 39.8803f,
+ 41.5188f, 55.2256f, 55.1803f, 55.6495f, 49.8920f, 34.8721f, 50.0859f, 57.0099f,
+ 47.3032f, 50.8975f, 47.4671f, 38.0605f, 41.4790f, 31.2737f, 42.9272f, 24.6954f,
+ 23.1770f, 22.9120f, 3.2977f, 35.6270f, 23.7935f, 12.0286f, 25.7104f, -2.4601f,
+ 06.7021f, 01.6804f, 02.0671f, -2.2891f, -16.2070f, -14.2204f, -20.1780f, -18.9303f,
+ -20.4859f, -25.8338f, -47.2892f
+};
+
+
+
+float32_t AData[NUMSAMPLES * NUMUNKNOWS];
+float32_t ATData[NUMSAMPLES * NUMUNKNOWS];
+float32_t ATAData[NUMUNKNOWS * NUMUNKNOWS];
+float32_t invATAData[NUMUNKNOWS * NUMUNKNOWS];
+float32_t BData[NUMSAMPLES * NUMSAMPLES];
+float32_t cData[NUMUNKNOWS];
+
+arm_matrix_instance_f32 t= {NUMSAMPLES, 1, tData};
+arm_matrix_instance_f32 x= {NUMSAMPLES, 1, xData};
+arm_matrix_instance_f32 A= {NUMSAMPLES, NUMUNKNOWS, AData};
+arm_matrix_instance_f32 AT= {NUMUNKNOWS, NUMSAMPLES, ATData};
+arm_matrix_instance_f32 ATA= {NUMUNKNOWS, NUMUNKNOWS, ATAData};
+arm_matrix_instance_f32 invATA= {NUMUNKNOWS, NUMUNKNOWS, invATAData};
+arm_matrix_instance_f32 B= {NUMUNKNOWS, NUMSAMPLES, BData};
+arm_matrix_instance_f32 c= {NUMUNKNOWS, 1, cData};
+
+main(void)
+{
+ int i, time;
+ float y;
+
+ pc.printf("/n/nMatrix calculation (Least squares) with 3 Unknowns \n\r");
+ pc.printf("We calculates a, b and c from; x(t)= a + b*t + c*t*t \n\r");
+ pc.printf("Results approx a= 8.701225, b= 38.880058, c= -9.792965\n\r");
+
+ y= sqrtf(xData[0]);
+ cData[0]= y;
+
+ for(i=0; i<NUMSAMPLES; i++) {
+ AData[i*NUMUNKNOWS +0] = 1.0f;
+ AData[i*NUMUNKNOWS +1] = tData[i];
+ AData[i*NUMUNKNOWS +2] = tData[i] * tData[i];
+ }
+
+ /* calculation of A transpose */
+ arm_mat_trans_f32(&A, &AT);
+ arm_mat_mult_f32(&AT, &A, &ATA);
+ arm_mat_inverse_f32(&ATA, &invATA);
+ arm_mat_mult_f32(&invATA, &AT, &B);
+ arm_mat_mult_f32(&B, &x, &c);
+
+
+ pc.printf("Results a= %f, b= %f, c= %f \n\r", cData[0], cData[1], cData[2]);
+
+ for(time=0; time<6; time++) {
+ y= cData[0] + cData[1]* time + cData[2] * time * time;
+ pc.printf("time= %d, x= %f \n\r", time, y );
+ }
+}
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/mbed.bld Tue Mar 10 12:14:31 2015 +0000 @@ -0,0 +1,1 @@ +http://mbed.org/users/mbed_official/code/mbed/builds/7e07b6fb45cf \ No newline at end of file