Skip to content

Data Model.md

Latest commit

 

History

History
265 lines (241 loc) · 15 KB

Data Model.md

File metadata and controls

265 lines (241 loc) · 15 KB

Data Model

Rather than holding onto individual fields, the concept of a fix is used to group data members of the GPS acquisition into a C structure (a struct type called gps_fix). This also facilitates merging pieces received at different times (i.e., in separate sentences) into a single easy-to-use structure.

The main NMEAGPS gps; object you declare in your sketch parses received characters, gradually assembling a fix. Most programs will call gps.read() to obtain the completed fix structure (see Usage below).

Given a variable declaration of type gps_fix:

gps_fix fix;

...this fix variable (or any other variable of type gps_fix) contains the following members:

  • fix.status, a status code
    • enum values STATUS_NONE, STATUS_EST, STATUS_TIME_ONLY, STATUS_STD or STATUS_DGPS
  • a location structure (i.e., latitude and longitude), accessed with
    • fix.latitudeL() and fix.longitudeL() for the higher-precision integer degrees, scaled by 10,000,000 (10 significant digits)
    • fix.latitude() and fix.longitude() for the lower-precision floating-point degrees (~7 significant digits)
    • NOTE: these lat/lon values are
      • positive for North or East degrees and negative for South or West degrees.
      • stored in a 'fix.location' structure, like a 2D coordinate. The location_t class provides additional methods for distance, bearing and offset calculations, as described here.
    • fix.latitudeDMS and fix.latitudeDMS are structures (see DMS.h) that each contain
      • fix.longitudeDMS.degrees in integer degrees
      • fix.latitudeDMS.degrees, in integer minutes
      • fix.longitudeDMS.seconds_whole, in integer seconds
      • fix.latitudeDMS.seconds_frac, in integer thousandths of a second
      • fix.latitudeDMS.secondsF(), in floating-point seconds
      • hemisphere indicator, accessed with
        • fix.longitudeDMS.hemisphere (enum values NORTH_H, SOUTH_H, EAST_H or WEST_H)
        • fix.longitudeDMS.EW() (char values E or W)
        • fix.latitudeDMS.NS() (char values N or S)
      • NOTE: An integer degree value (scaled by 107 can be used to set the DMS structure by using fix.latitudeDMS.From( otherLatitude );
  • an altitude (above ellipsoid, aka Mean Sea Level), accessed with
    • fix.altitude_cm(), in integer centimeters
    • fix.altitude(), in floating-point meters
    • fix.alt.whole, in integer meters
    • fix.alt.frac, in integer centimeters, to be added to the whole part
  • a speed, accessed with
    • fix.speed_kph(), in floating-point kilometers per hour
    • fix.speed_mph(), in floating-point miles per hour
    • fix.speed(), in floating-point knots (nautical miles per hour)
    • fix.speed_mkn(), in integer knots, scaled by 1000
    • fix.spd.whole, in integer knots
    • fix.spd.frac, in integer thousandths of a knot, to be added to the whole part
  • a heading, accessed with
    • fix.heading_cd(), in integer hundredths of a degree
    • fix.heading(), in floating-point degrees
  • velocity components in the North, East and Down directions, accessed with
    • fix.velocity_north, in integer cm/s
    • fix.velocity_east, in integer cm/s
    • fix.velocity_down, in integer cm/s
  • fix.hdop, fix.vdop and fix.pdop, in integer thousandths of the DOP.
    • Dilution of Precision is a unitless measure of the current satellite constellation geometry WRT how 'good' it is for determining a position. This is independent of signal strength and many other factors that may be internal to the receiver.   It cannot be used to determine position accuracy in meters. Instead, use the LAT/LON/ALT error in cm members, which are populated by GST sentences.
  • latitude, longitude and altitude error, accessed with
    • fix.lat_err_cm, fix.lon_err_cm and fix.alt_err_cm, in integer centimeters
    • fix.lat_err(), fix.lon_err() and fix.alt_err(), in floating-point meters
  • speed, heading and time errors, accessed with
    • fix.spd_err_mmps, in integer mm/s
    • fix.hdg_errE5, in integer degrees * 100000
    • fix.time_err_ns, in integer nanoseconds
      or with
    • fix.spd_err() in floating-point m/s
    • fix.hdg_err() in floating-point degrees
    • fix.time_err() in floating-point seconds
  • geoid height above ellipsoid (see here for description), accessed with
    • fix.geoidHeight_cm, in integer centimeters
    • fix.geoidHeight(), in floating-point meters
    • fix.geoidHt.whole, in integer meters
    • fix.geoidHt.frac, in integer centimeters to be added to the whole part
  • fix.satellites, a satellite count
  • a date/time structure (see Time.h), accessed with
    • fix.dateTime.year,
    • fix.dateTime.month,
    • fix.dateTime.date, the day-of-month,
    • fix.dateTime.hours,
    • fix.dateTime.minutes,
    • fix.dateTime.seconds, and
    • fix.dateTime.day, the day-of-the-week. This member is expensive to calculate, so it is uninitialized until you call the set_day() method. If you need the day-of-the-week, be sure to call set_day whenever the year, month or date members are changed. In general, call fix.dateTime.set_day() whenever fix is assigned (e.g., fix = gps.read()).

      Time operations allow converting to and from total seconds offset from a de facto starting time (e.g., an epoch date/time "origin"). There are constants in Time.h for NTP, POSIX and Y2K epochs. Simply change the static members s_epoch_year and s_epoch_weekday in Time.h, and all date/time operations will be based on that epoch. This does not affect GPS times, but it will allow you to easily convert a GPS time to/from an NTP or POSIX time value (seconds).

      The NMEAtimezone.ino example program shows how to convert the GPS time (UTC) into a local time. Basically, a Time structure is converted to seconds (from the epoch start), then the time zone offset in seconds is added, and then the offset seconds are converted back to a time structure, with corrected day, month, year, hours and minutes members.
  • fix.dateTime_cs, in integer hundredths of a second
    • fix.dateTime_ms(), in milliseconds
    • fix.dateTime_us(), in microseconds
  • a collection of boolean valid flags for each of the above members, accessed with
    • fix.valid.status
    • fix.valid.date for year, month, day-of-month
    • fix.valid.time for hours, minutes, seconds and centiseconds
    • fix.valid.location for latitude and longitude
    • fix.valid.altitude
    • fix.valid.speed
    • fix.valid.heading
    • fix.valid.hdop, fix.valid.vdop and fix.valid.pdop
    • fix.valid.lat_err, fix.valid.lon_err and fix.valid.alt_err
    • fix.valid.geoidHeight

Validity

Because the GPS device may not have a fix, each member of a gps_fix can be marked as valid or invalid. That is, the GPS device may not know the lat/long yet. To check whether the fix member has been received, test the corresponding valid flag (described above). For example, to check if lat/long data has been received:

  if (my_fix.valid.location) {
    Serial.print( my_fix.latitude() );
    Serial.print( ',' );
    Serial.println( my_fix.longitude() );
  }

You should also know that, even though you have enabled a particular member (see GPSfix_cfg.h), it may not have a value until the related NMEA sentence sets it. And if you have not enabled that sentence for parsing in NMEAGPS_cfg.h, it will never be valid.

Other GPS-related information

There is additional information that is not related to a fix. Instead, it contains information about parsing or a Global Navigation Satellite System. GNSS's currently include GPS (US), GLONASS (Russia), Beidou (China) and Galileo (EU). The main NMEAGPS gps object you declare in your sketch contains:

  • gps.UTCsecondStart(), the Arduino micros() value when the current UTC second started
  • gps.UTCms(), the number of milliseconds since the last received UTC time, calculated from micros() and gps.UTCsecondStart.
  • gps.UTCus(), the number of microseconds since the last received UTC time, calculated from micros() and gps.UTCsecondStart.
  • gps.nmeaMessage, the latest received message type. This is an ephemeral value, because multiple sentences are merged into one fix structure. If you only check this after a complete fix is received, you will only see the LAST_SENTENCE_IN_INTERVAL.
    • enum values NMEA_GLL, NMEA_GSA, NMEA_GST, NMEA_GSV, NMEA_RMC, NMEA_VTG or NMEA_ZDA
  • gps.satellies[], an array of satellite-specific information, where each element contains
    • gps.satellies[i].id, satellite ID
    • gps.satellies[i].elevation, satellite elevation in 0-90 integer degrees
    • gps.satellies[i].azimuth, satellite azimuth in 0-359 integer degrees
    • gps.satellies[i].snr, satellite signal-to-noise ratio in 0-99 integer dBHz
    • gps.satellies[i].tracked, satellite being tracked flag, a boolean
  • gps.sat_count, the number of elements in the gps.satellites[] array
  • gps.talker_id[], talker ID, a two-character array (not NUL-terminated)
  • gps.mfr_id[], manufacturer ID, a three-character array (not NUL-terminated)
  • an internal fix structure, gps.fix(). Most sketches should not use gps.fix() directly!

Usage

First, declare an instance of NMEAGPS:

NMEAGPS gps;

Next, tell the gps object to handle any available characters on the serial port:

void loop()
{
  while (gps.available( gps_port )) {

The gps object will check if there are any characters available, and if so, read them from the port and parse them into its internal fix. Many characters will have to be read before the current fix is complete, so gps.available will return false until the fix is complete; the body of while loop will be skipped many times, and the rest of loop() will be executed.

When a fix is finally completed, gps.available will return true. Now your sketch can "read" the completed fix structure from the gps object:

void loop()
{
  while (gps.available( gps_port )) {
    gps_fix fix = gps.read();

The local fix variable now contains all the GPS fields that were parsed from the gps_port. You can access them as described above:

void loop()
{
  while (gps.available( gps_port )) {
    gps_fix fix = gps.read();
    if (fix.valid.time) {
       ...

Note that the fix variable is local to that while loop; it cannot be accessed elsewhere in your sketch. If you need to access the fix information elsewhere, you must declare a global fix variable:

gps_fix currentFix;

void loop()
{
  while (gps.available( gps_port )) {
    currentFix = gps.read();
    if (currentFix.valid.time) {
       ...

Any part of your sketch can use the information in currentFix.

Please note that the fix structure is much smaller than the raw character data (sentences). A fix is nominally 1/4 the size of one sentence (~30 bytes vs ~120 bytes). If two sentences are sent during each update interval, a fix could be 1/8 the size required for buffering two sentences.

In this fix-oriented program structure, the methods gps.available and gps.read are manipulating entire gps_fix structures. Multiple characters and sentences are used internally to fill out a single fix: members are "merged" from sentences into one fix structure (described here).

That program structure is very similar to the typical serial port reading loop:

void loop()
{
  while (serial.available()) {
    char c = serial.read();
       ... do something with the character ...;
  }

However, the fix-oriented methods operate on complete fixes, not individual characters, fields or sentences.

Note: If you find that you need to filter or merge data with a finer level of control, you may need to use a different Merging option, Coherency, or the more-advanced Character-Oriented methods.

Examples

Some examples of accessing fix values:

gps_fix fix_copy = gps.read();

int32_t lat_10e7 = fix_copy.lat; // scaled integer value of latitude
float lat = fix_copy.latitude(); // float value of latitude

Serial.print( fix_copy.latDMS.degrees );
Serial.print( ' ' );
Serial.print( fix_copy.latDMS.minutes );
Serial.print( F("' " );
Serial.print( fix_copy.latDMS.seconds );

if (fix_copy.dateTime.month == 4) // test for the cruelest month
  cry();

// Count how satellites are being received for each GNSS
for (uint8_t i=0; i < gps.sat_count; i++) {
  if (gps.satellites[i].tracked) {
    if (gps.satellites[i] . id <= 32)
      GPS_satellites++;
    if (gps.satellites[i] . id <= 64)
      SBAS_satellites++;
    if (gps.satellites[i] . id <= 96)
      GLONASS_satellites++;
  }
}

And some examples of accessing valid flags in a fix structure:

if (fix_copy.valid.location)
  // we have a lat/long!

if (fix_copy.valid.time)
  // the copy has hours, minutes and seconds

Here's an example for accessing the altitude

    if (fix_copy.valid.altitude) {
      z2 = fix_copy.altitude_cm();
      vz = (z2 - z1) / dt;
      z1 = z2;

      // Note: if you only care about meters, you could also do this:
      //    z = fix_copy.alt.whole;
    }

You can also check a collection of flags before performing a calculation involving multiple members:

    if (fix_copy.valid.altitude && fix_copy.valid.date && fix_copy.valid.time) {
      dt = (clock_t) fix_copy.dateTime - (clock_t) fix_copy.dateTime;
      dz = fix_copy.alt.whole - last_alt;  // meters
      vz = dz / dt;                         // meters per second vertical velocity
    }

Bonus: The compiler will optimize this into a single bit mask operation.

The example printing utility file, Streamers.cpp shows how to access each fix member and print its value.

Options

Except for status, each of these gps_fix members is conditionally compiled; any, all, or no members can be selected for parsing, storing and merging. This allows you to configuring NeoGPS to use the minimum amount of RAM for the particular members of interest. See Configurations for how to edit GPSfix_cfg.h and NMEAGPS_cfg.h, respectively.

Precision

Integers are used for all members, retaining full precision of the original data.

    gps_fix fix = gps.read();
    if (fix.valid.location) {
      // 32-bit ints have 10 significant digits, so you can detect very
      // small changes in position:
      d_lat = fix_copy.lat - last_lat;
    }

Optional floating-point accessors are provided for many members.

    if (fix_copy.valid.location) {
      float lat = fix_copy.latitude();

      // floats only have about 6 significant digits, so this
      // computation is useless for detecting small movements:
      foobar = (lat - target_lat);