In order to develop the final integrated fire control system, the number of off- boresight commands which can/will be used to shape the weapon trajectory must be established. The first off-boresight command is programmed into the weapon while on the launcher at T=0s. Subsequent commands will be transmitted to the weapon over the data link. The time steps at which the off-boresight commands will be transmitted will depend on the information obtained about the target from the tracking and prediction system. The appropriate number of off-boresight commands which should be used to shape the weapon trajectory over the flight period, can be established by considering the fire control system as operating in two phases during a simulated engagement. These phases consist of:
1. Weapon initialisation at T=0s.
2. In-flight trajectory revision.
At T=0s, the target will have entered the exclusion zone i.e. it will lie within the reachable set of the weapon. Any vessel which would have approached that exclusion zone would be warned to leave. Therefore, the logical assumption will be made that the target will have deliberately entered that zone and is seeking to attack the ship [38]. The trajectory of the weapon should therefore be shaped considering an attack by the target. As discussed by [35], in an attack, the target will want to minimise the time taken to reach the ship. In the most simple case it would favour a direct attack but it may have to manoeuvre for a number of reasons such as to achieve a specific attack heading.
The probabilities of the predicted target distribution which will be defined at this time step will initially weight the probability of a non manoeuvring target trajectory higher than the other trajectories thereby reflecting this assumption.
From Chapter 3, it is apparent that the seeker of the weapon should be on for 35s of flight, therefore the integrated fire control system should seek to shape the trajectory of the weapon such that the target should be detected at the latest in the 30-35s time interval.
The integrated tracking and prediction system will provide 6 updates in this time period. If the target performs a direct attack, then these updates (which will include an updated distribution of possible target trajectories and associ- ated probabilities) will occur every 5s. It could be concluded that the trajectory should be therefore be shaped based on 7 off-boresight commands, which would initially be transmitted to the weapon at 5s intervals to coincide with the tracking and prediction system updates. However revising the trajectory at this update rate if the target has not manoeuvred could be inefficient in terms of the cost to the maximum heading change which can be applied at subsequent off-boresight commands.
Each off-boresight command which is not equal to the previous off-boresight requires a heading change. Though the new off-boresight heading will not be achieved instantly, the size of the maximum heading change which can be used in subsequent commands is reduced instantly within the fire control system.
If the target has not manoeuvred then the change in the probability distribu- tion will be only be minor, especially in the early stages of the engagement. It is therefore unlikely that the optimal trajectory would need to be significantly revised at 5s time intervals. The maximum heading change capability would be better preserved in case the target does perform a manoeuvre. Therefore 4 off- boresight commands will be initially used to shape the trajectory of the weapon which are calculated considering that they will be transmitted to the weapon at T=0s, T=10s, T=20s and T=30s.
It is expected that the intercept probability distribution which will be associ- ated with these trajectory constraints will have a large number of local maxima based on Figures 5.3 and 5.4. The simulated annealing algorithm is therefore used at the initialisation stage to determine the optimal weapon trajectory based
on an assumed complicated intercept probability distribution.
Once the weapon has left the launcher, the system will then enter the in-flight trajectory revision phase. The process by which the trajectory of the weapon is revised during this phase depends on how the target behaves over the flight time of the weapon. If the target obeys the assumption of a direct attack, then the trajectory will be revised at T=10s and T=20s using the simulated annealing process. If the weapon seeker has not detected the target prior to T=30s, then the trajectory at T=30s will be fine-tuned using a simple search algorithm. The simple search algorithm will seek to determine the appropriate off-boresight com- mand to be applied at T=30s, considering only a very small number of remaining possible target trajectories.
If the target manoeuvres during the weapon fly out, then the in-flight trajec- tory revision phase will optimise the trajectory in a variety of ways, depending on the duration and number of the manoeuvres, using both the simulated anneal- ing and simple search algorithms.
The initial calculation of the optimal trajectory and subsequent revision due to different target behaviours will be discussed in detail in Sections 5.8 and 5.9 respectively. However the integration of the weapon data link and the trajectory and prediction system is discussed first in order to aid the reader’s understanding of the whole system in later sections.
5.7.1
Data Link and Tracking and Prediction System In-
tegration
The data link is assumed to be an on-demand service. This means that the weapon could technically receive any number of off-boresight commands over the fly out. The off-boresight commands could be transmitted at any given point in time. However, as the weapon is to be deployed within an integrated fire con- trol system, the data link capability is restricted. New off-boresight commands will only be transmitted by the fire control system if an optimisation process has been performed. An optimisation process will be performed based on the up- dated target trajectory prediction. The off-boresight commands will therefore be transmitted at the same time step that the tracking and prediction system pro- duces updated target behaviour and an optimisation process has been performed.
The time taken for the optimisation process algorithm to run is neglected. The assumption is made if the system was deployed operationally, sufficient comput-
ing power would be available to complete the optimisation process in real time.
As the tracking system can only make one manoeuvre detection in a 5s period or update considering constant velocity behaviour, the fire control system will only be able to transmit one off-boresight command to the weapon in each 5s period. As the IMM has an average settling time of 1.4s, off-boresight commands will only be transmitted after the first 1.4s of each state transition period.