|
Techniques and tools for validation of real-time infrared target signature models are presented. The model validation techniques presented in this paper were developed for hardware-in-the-loop (HWIL) simulations at the U.S. Army Missile Command's Research, Development, and Engineering Center. Real-time target model validation is a required deliverable to the customer of a HWIL simulation facility and is a critical part of ensuring the fidelity of a HWIL simulation. There are two levels of real-time target model validation. The first level is comparison of the target model to some baseline or measured data which answers the question `are the simulation inputs correct?'. The second level of validation is a simulation validation which answers the question `for a given target model input is the simulation hardware and software generating the correct output?'. This paper deals primarily with the first level of target model validation. IR target signature models have often been validated by subjective visual inspection or by objective, but limited, statistical comparisons. Subjective methods can be very satisfying to the simulation developer but offer little comfort to the simulation customer since subjective methods cannot be documented. Generic statistical methods offer a level of documentation, yet are often not robust enough to fully test the fidelity of an IR signature. Advances in infrared seeker and sensor technology have led to the necessity of system specific target model validation. For any HWIL simulation it must be demonstrated that the sensor responds to the real-time signature model in a manner which is functionally equivalent to the sensor's response to a baseline model. Depending on the application, a baseline method can be measured IR imagery or the output of a validated IR signature prediction code. Tools are described that generate validation data for HWIL simulations at MICOM and example real-time model validations are presented.
|
