What happened to meaRng?

jstroup1986 · November 23, 2021, 12:35am

Going through the setup process of my CubePilot HereFlow, I came across reference to a parameter in the docs that doesn’t seem to exist any more: meaRng

The docs reference evaluating EKF5.meaRng. EKF5 doesn’t exist any more - an after looking through every single parameter in the dataflash logs, meaRng doesn’t seem to exist any more either.

From the references I’ve seen, this parameter simply logs the distances the HereFlow Lidar rangefinder is registering over time. But I’m not sure.

I plan to install and use a Lightware LW20/C as a rangefinder - but I want to know exactly what the HereFlow is doing before making the change.

There are quite a few posts inquiring about this over the past few years - but no answers.

Thank you for any assistance anyone can offer.

dkemxr · November 23, 2021, 1:26am

I’ll take a guess and suggest NKF5.rng

jstroup1986 · November 23, 2021, 1:57am

Thanks Dave,

I think someone else suggested that too.

Unfortunately, the BIN files for my test flights with a HereFlow installed and the lidar RangeFinder enabled - don’t have NKF5 parameters available to graph.

I know the HereFlow’s lidar is working - because it’s measurements show up in the “sonarrange” on the Status tab on Mission Planner.

Seems to me there’s got to be something Mission Planner can graph with lidar measurements. But maybe it got left out in a recent revision.

dkemxr · November 23, 2021, 2:03am

…

XKF5.rng then. I happened to be looking at a V4.0.7 log when I suggested NKF.

jstroup1986 · November 23, 2021, 2:10am

Looks like it. This is the graph of XKF5.mg from one of my X and Y axis calibration test for the HereFlow:

This is consistent with what would be expected.

Interesting that the independent axis for this graph is labeled “unknown.”

Thanks Dave!

peterbarker · November 26, 2021, 4:01am

That’s down to one of the question marks on this line:

github.com

ardupilot/ardupilot/blob/master/libraries/AP_NavEKF3/LogStructure.h#L435

    
      
            "XKF0","QBBccCCcccccccc","TimeUS,C,ID,rng,innov,SIV,TR,BPN,BPE,BPD,OFH,OFL,OFN,OFE,OFD", "s#-m---mmmmmmmm", "F--B---BBBBBBBB" , true }, \
          { LOG_XKF1_MSG, sizeof(log_XKF1), \
            "XKF1","QBccCfffffffccce","TimeUS,C,Roll,Pitch,Yaw,VN,VE,VD,dPD,PN,PE,PD,GX,GY,GZ,OH", "s#ddhnnnnmmmkkkm", "F-BBB0000000BBBB" , true }, \
          { LOG_XKF2_MSG, sizeof(log_XKF2), \
            "XKF2","QBccccchhhhhhfff","TimeUS,C,AX,AY,AZ,VWN,VWE,MN,ME,MD,MX,MY,MZ,IDX,IDY,IS", "s#---nnGGGGGGoor", "F----BBCCCCCC000" , true }, \
          { LOG_XKF3_MSG, sizeof(log_XKF3), \
            "XKF3","QBcccccchhhccff","TimeUS,C,IVN,IVE,IVD,IPN,IPE,IPD,IMX,IMY,IMZ,IYAW,IVT,RErr,ErSc", "s#nnnmmmGGGd?--", "F-BBBBBBCCCBB00" , true }, \
          { LOG_XKF4_MSG, sizeof(log_XKF4), \
            "XKF4","QBcccccfffHBIHb","TimeUS,C,SV,SP,SH,SM,SVT,errRP,OFN,OFE,FS,TS,SS,GPS,PI", "s#------mm-----", "F-------??-----" , true }, \
          { LOG_XKF5_MSG, sizeof(log_XKF5), \
            "XKF5","QBBhhhcccCCfff","TimeUS,C,NI,FIX,FIY,AFI,HAGL,offset,RI,rng,Herr,eAng,eVel,ePos", "s#----m???mrnm", "F-----BBBBB000" , true }, \
          { LOG_XKFD_MSG, sizeof(log_XKFD), \
            "XKFD","QBffffff","TimeUS,C,IX,IY,IZ,IVX,IVY,IVZ", "s#------", "F-------" , true }, \
          { LOG_XKFM_MSG, sizeof(log_XKFM),   \
            "XKFM", "QBBffff", "TimeUS,C,OGNM,GLR,ALR,GDR,ADR", "s#-----", "F------", true }, \
          { LOG_XKFS_MSG, sizeof(log_XKFS), \
            "XKFS","QBBBBBB","TimeUS,C,MI,BI,GI,AI,SS", "s#-----", "F------" , true }, \
          { LOG_XKQ_MSG, sizeof(log_XKQ), "XKQ", "QBffff", "TimeUS,C,Q1,Q2,Q3,Q4", "s#????", "F-????" , true }, \
          { LOG_XKT_MSG, sizeof(log_XKT),   \
            "XKT", "QBIffffffff", "TimeUS,C,Cnt,IMUMin,IMUMax,EKFMin,EKFMax,AngMin,AngMax,VMin,VMax", "s#sssssssss", "F-000000000", true }, \
          { LOG_XKTV_MSG, sizeof(log_XKTV),                         \

When I put in the initiali logging documentation, if I wasn’t sure of a
unit and couldn’t work it out immediately I put in a “?” so it could be
addressed later.

It’s a bit odd that there is a multiplier in there.

Nobody has gotten back to actually verifying that this is in metres or
centimetres in the logfile

jstroup1986 · November 26, 2021, 12:12pm

While I discovered this parameter when setting up my HereFlow that has a lidar limited to 200cm, I’ve since added a Lightware LW20/C with a lidar limit of 9500cm.

Since the lidar on the HereFlow is typically disabled, I can see why this parameter has received little attention. But it also reports the Lightware LW20/C’s data - which is much more relevant. Odd that the docs don’t mention this parameter in the section about the Lightware lidar sensors.

Since multiple rangefinders are supported - what happens when multiple are installed and configured?
I plotted out XKF5.mg and baro altitude from a mission I flew yesterday. It does seem like the independent axis is meters.

It’s interesting to see how the XKF5.mg levels off at 95 meters - rngfnd1_max_cm set to “9500” - so the log retains that highest setting.

The docs note that the rangefinder can be disabled with an aux-switch channel - in case erroneous data causes a problem. As this mission was set to fly at 400’ - so the lidar data is wrong above 9500cm. I need to figure out if this is an issue.

hendjosh · December 6, 2021, 1:09am

The odd multiplier is because the type is because of the (int16_t)(100 x zzzz).

Thoughts on making those all floats instead? Downside of course being double the space…

github.com

ArduPilot/ardupilot/blob/522056d6f79bc089dfdcff93bb4161830cb2511e/libraries/AP_NavEKF3/AP_NavEKF3_Logging.cpp#L185-L203

    
      
              const struct log_XKF5 pkt5{
                  LOG_PACKET_HEADER_INIT(LOG_XKF5_MSG),
                  time_us : time_us,
                  core    : DAL_CORE(core_index),
                  normInnov : (uint8_t)(MIN(100*MAX(flowTestRatio[0],flowTestRatio[1]),255)),  // normalised innovation variance ratio for optical flow observations fused by the main nav filter
                  FIX : (int16_t)(1000*flowInnov[0]),  // optical flow LOS rate vector innovations from the main nav filter
                  FIY : (int16_t)(1000*flowInnov[1]),  // optical flow LOS rate vector innovations from the main nav filter
                  AFI : (int16_t)(1000 * auxFlowObsInnov.length()),  // optical flow LOS rate innovation from terrain offset estimator
                  HAGL : (int16_t)(100*(terrainState - stateStruct.position.z)),    // height above ground level
                  offset : (int16_t)(100*terrainState),           // filter ground offset state error
                  RI : (int16_t)(100*innovRng),                   // range finder innovations
                  meaRng : (uint16_t)(100*rangeDataDelayed.rng),  // measured range
                  errHAGL : (uint16_t)(100*sqrtF(Popt)),          // note Popt is constrained to be non-negative in EstimateTerrainOffset()
                  angErr : (float)outputTrackError.x,             // output predictor angle error
                  velErr : (float)outputTrackError.y,             // output predictor velocity error
                  posErr : (float)outputTrackError.z              // output predictor position tracking error
              };
              AP::logger().WriteBlock(&pkt5, sizeof(pkt5));
          }