forked from bitcraze/crazyflie-firmware
-
Notifications
You must be signed in to change notification settings - Fork 1
/
tfmicrodemo.c
263 lines (236 loc) · 7.65 KB
/
tfmicrodemo.c
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
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
/* TFMicro Test Script: Tests to see the delay incurred by TF-Micro and
how fast the chip can process these neural networks */
#include "deck.h"
#include "system.h"
#include "commander.h"
#include "range.h" // get the 6axis distance measurements
#include "FreeRTOS.h"
#include "task.h"
#include "debug.h"
#include "machinelearning.h"
#include "sysload.h"
#include "sequencelib.h"
#include "sensor.h"
#define TENSOR_ALLOC_SIZE 6000
#define SUBTRACT_VAL 60
#define STATE_LEN 5
#define NUM_STATES 4
#define YAW_INCR 8
//#define SENS_MIN 35000
#define SENS_MIN 11000
#define SENS_MAX 55000
#define TRUE 1
#define FALSE 0
#define GOAL_THRES 245
#define GOAL_THRES_COUNT 3
#define DIST_MIN 90
#define RAND_ACTION_RATE 30
void yaw_incr(int *yaw){
int yaw_out = *yaw + YAW_INCR;
if(yaw_out>180){
yaw_out -= 360;
}
*yaw = yaw_out;
return;
}
void yaw_decr(int *yaw){
int yaw_out = *yaw - YAW_INCR;
if(yaw_out<-180){
yaw_out += 360;
}
*yaw = yaw_out;
return;
}
static void check_multiranger_online() {
DEBUG_PRINT("Checking if multiranger ToF sensors online...\n");
distances d;
for (int j = 0; j < 10; j++) {
getDistances(&d);
vTaskDelay(M2T(100));
}
if (d.left == 0 && d.right == 0) {
// most likely the ranger deck isn't attached correctly
DEBUG_PRINT("Most likely ranger deck not attached correctly\n");
for (;;) {
vTaskDelay(M2T(1000));
}
}
}
uint8_t get_distance(uint16_t sensor_read){
if(sensor_read<SENS_MIN){
sensor_read = SENS_MIN;
}
if(sensor_read>SENS_MAX){
sensor_read=SENS_MAX;
}
float frac = ((float)sensor_read-SENS_MIN)/(SENS_MAX-SENS_MIN) ;
return (uint8_t)(frac*255);
}
static void update_state(uint8_t *meas_array, distances d,uint8_t dist){
//Step 1: move entire array by 1 state
for(int i = (STATE_LEN*NUM_STATES-1);i>=STATE_LEN;i--)
{
*(meas_array+i) = *(meas_array+i-STATE_LEN);
}
//Step2: update the first state
*(meas_array) = (uint8_t) ( d.right * 0.06375);
*(meas_array+1) = (uint8_t) ( d.front * 0.06375);
*(meas_array+2) = (uint8_t) ( d.left * 0.06375);
*(meas_array+3) = (uint8_t) ( d.back * 0.06375);
*(meas_array+4) = (uint8_t) (dist);
}
static void tfMicroDemoTask()
{
static setpoint_t setpoint;
systemWaitStart();
const CTfLiteModel* model = CTfLiteModel_create();
uint8_t tensor_alloc[TENSOR_ALLOC_SIZE];
int r[3];
uint8_t full_meas[20] = {40, 120, 120, 120, 120, 40, 120, 120, 120, 120, 40, 120, 120, 120, 120, 40, 120, 120, 120, 120};
uint8_t input[5] = {40, 120, 120, 120, 120};
uint16_t sensor_read = 0;
uint8_t sensor_mode = 0;
DEBUG_PRINT("Starting the advanced machine learning...\n");
float HOVER_HEIGHT = 0.8;
// Start in the air before doing ML
flyVerticalInterpolated(0.0f, HOVER_HEIGHT, 6000.0f);
vTaskDelay(M2T(500));
distances d;
getDistances(&d);
TSL2591_init();
uint8_t rand_arr[10] = {2, 2, 1, 1, 2, 1, 1, 1, 2, 1};
uint8_t dist =0;
int yaw = 0;
int command = 0;
float ESCAPE_SPEED = 0.3;
uint8_t goal_count = 0;
uint8_t found_goal = FALSE;
uint8_t rand_count = 0;
for (int j = 0; j < 10000; j++) {
getDistances(&d);
//DEBUG_PRINT("yaw: %d \n",yaw);
/* Defining the input to the network*/
// obs avoidance will
// input[0] = (uint8_t) ( d.front / 10);
// input[1] = (uint8_t) ( d.right / 10);
// input[2] = (uint8_t) ( d.back / 10);
// input[3] = (uint8_t) ( d.left / 10);
// input[4] = (uint8_t) ( d.up / 10);
// input[5] = (uint8_t) ( d.down / 10);
if(d.up/10 < 20)
{
break;
}
// if(dist>GOAL_THRES){
// goal_count++;
// if(goal_count>=GOAL_THRES_COUNT){
// break;
// }
// }
// DEBUG_PRINT("%i \n",sensor_read);
// dist = 128;
// vTaskDelay(M2T(500));
sensor_read = read_TSL2591(sensor_mode);
dist = get_distance(sensor_read);
//vTaskDelay(M2T(300));
//DEBUG_PRINT("FRONT : %f\n",(float)(d.front)*0.001);
input[0] = (uint8_t) ( d.right* 0.06375);
input[1] = (uint8_t) ( d.front * 0.06375);
input[2] = (uint8_t) ( d.left * 0.06375);
input[3] = (uint8_t) ( d.back * 0.06375);
input[4] = (uint8_t) dist;
update_state(&full_meas, d,dist);
//DEBUG_PRINT("full meas: %i %i %i %i %i %i %i %i %i %i %i %i %i %i %i %i %i %i %i %i \n",full_meas[0],full_meas[1],full_meas[2],full_meas[3],full_meas[4],full_meas[5],full_meas[6],full_meas[7],
// full_meas[8],full_meas[9], full_meas[10],full_meas[11],full_meas[12],full_meas[13],full_meas[14],full_meas[15],full_meas[16],full_meas[17],full_meas[18],full_meas[19]);
DEBUG_PRINT("sensor %i \n",sensor_read);
DEBUG_PRINT("dist %i \n",dist);
// subtract from laser readings, this creates a save zone around objects
// for(int i=0;i<4;i++)
// {
// if(input[i]>SUBTRACT_VAL){
// input[i] = input[i] - SUBTRACT_VAL;
// }
// else{
// input[i] = 0;
// }
// }
// DEBUG_PRINT("LASERS: %i %i %i %i \n",input[0],input[1],input[2],input[3]);
// input[0] = (uint8_t)(1);
// input[1] = (uint8_t)(1);
// input[2] = (uint8_t)(1);
// input[3] = (uint8_t)(1);
// input[4] = (uint8_t)(1);
CTfInterpreter_simple_fc(model, tensor_alloc, TENSOR_ALLOC_SIZE, full_meas, r);
//DEBUG_PRINT("Q-Vals: %i %i %i \n",r[0],r[1],r[2]);
command = argmax(r, 3);
// if(j%RAND_ACTION_RATE==0)
// {
// if(command != 0){
// if(command ==1)
// {
// command = 2;
// }
// else{
// command = 1;
// }
// }
// else{
// command = rand_arr[rand_count%10];
// rand_count++;
// }
//
// }
//DEBUG_PRINT("Command: %i\n", command);
switch (command) {
case 0:
// setHoverSetpoint(&setpoint, ESCAPE_SPEED, 0, HOVER_HEIGHT, (float)(yaw));
setHoverSetpoint(&setpoint, ESCAPE_SPEED, 0, HOVER_HEIGHT, 0);
commanderSetSetpoint(&setpoint, 3);
vTaskDelay(M2T(150));
break;
case 1:
// yaw_incr(&yaw);
// setHoverSetpoint(&setpoint, 0, 0, HOVER_HEIGHT,(float)(yaw));
setHoverSetpoint(&setpoint, 0, 0, HOVER_HEIGHT, 54);
commanderSetSetpoint(&setpoint, 3);
vTaskDelay(M2T(150));
// vTaskDelay(M2T(100));
break;
case 2:
// yaw_decr(&yaw);
// setHoverSetpoint(&setpoint, 0, 0, HOVER_HEIGHT, (float)(yaw));
setHoverSetpoint(&setpoint, 0, 0, HOVER_HEIGHT, -54);
commanderSetSetpoint(&setpoint, 3);
vTaskDelay(M2T(150));
// vTaskDelay(M2T(100));
break;
default:
// setHoverSetpoint(&setpoint, 0, 0, HOVER_HEIGHT, (float)(yaw));
setHoverSetpoint(&setpoint, 0, 0, HOVER_HEIGHT, 0);
commanderSetSetpoint(&setpoint, 3);
vTaskDelay(M2T(40));
break;
}
//
}
// Slowly lower to a safe height before quitting, or else CRASH!
flyVerticalInterpolated(HOVER_HEIGHT, 0.1f, 1000.0f);
for (;;) { vTaskDelay(M2T(1000)); }
}
static void init() {
xTaskCreate(tfMicroDemoTask, "tfMicroDemoTask",
4500 /* Stack size in terms of WORDS (usually 4 bytes) */,
NULL, /*priority*/3, NULL);
}
static bool test() {
return true;
}
const DeckDriver tf_micro_demo = {
.vid = 0,
.pid = 0,
.name = "tfMicroDemo",
.usedGpio = 0, // FIXME: set the used pins
.init = init,
.test = test,
};
DECK_DRIVER(tf_micro_demo);