TABLE OF CONTENTS First, we shall generate two fight trajectories: one trajectory for an intended straight line path and other actual trajectory perturbed by small accidental accelerations in both x and y (see Figure 8.3).
FLT_XY.CPP generates the ideal path, TRAJ_XY.CPP generates an actual path. The acceleration perturbation is assumed distributed uniformly with zero mean and variance ? x 2 = ? y 2 = 8.0 (m/sec) 2.
The measurement data is generated in MEA_XY.CPP. The measurement errors in range and angle are also assumed distributed uniformly with zero mean and variance ? 2 r = 1,850 m 2, ? 2 ? = 1.015e ?4 rad 2.
Kalman filter processing is programmed in KAL_XY.CPP with a header file MATRIX.H. All the relevant data computed is recorded. The data is listed as follows:
| 1. Pest.DAT: | Estimated error covariance matrix; |
| 2. Ppre.DAT: | Predicted error covariance matrix; |
| 3. K_gain.DAT: | Kalman gain matrix; |
| 4. Xest.DAT: | Estimated state (estimated position); |
| 5. Xpre.DAT: | Predicted state (predicted position). |
We have generated two additional data files from (4) and (5) above for a post-fight analysis. They are listed as follows:
| 6. EST_RAZ.CPP: | Estimated range and azimuth angle; |
| 7. PRE_RAZ.CPP: | Predicted range and azimuth angle. |
The last two data files plus FLT_XY.DAT, TRAJ_XY.CPP and MEA_RAZ.DAT are used for a postflight analysis and graphic displays, Figures 8.9 and 8.10.
We notice that...
