This Small Business Innovation Research Phase I project deals with a new methodology for real-time testing of embedded controllers and other actual system hardware using simulation of major parts of the system. For example, a jet engine controller might be interfaced to the real-time simulation of a jet engine to test the engine controller over the full operating envelop of the engine. Simulations of dynamic systems involving `hard-in-the-loop` are commonly called HIL simulations. This new methodology involves completely asynchronous operation of processors and hardware subsystems and has the potential for producing enormous simplifications and gains in efficiency for HIL simulations. In this new methodology, variable step-size integration methods are used for the first time in hard real-time HIL simulations. This is a major breakthrough for this type of simulation. The extrapolation techniques used in the asynchronous methodology compensate for latencies inherent in a simulation having multiple processors and external hardware, thus providing greater simulation accuracy. The basic concepts of the asynchronous methodology can now be demonstrated in a multi-processor environment. A commercial software package containing a development shell, run-time system, and debug/monitor package that incorporates this new asynchronous methodology is envisioned. The primary output for this research would be a specification for such a software package. An example simulation with external hardware interfaced to it will be developed and used to test the asynchronous methodology concepts and in experiments that help define various aspects of the proposed software package, particularly in the areas of debugging tools and the interpretation of calculated estimates of dynamic errors in the simulation. The use of embedded microprocessor-based controllers for all types of products has literally exploded over the past few years. HIL simulation is a proven tool for reducing time to market for new products, improving product quality, and reducing the potential for product liability claims by making it possible to test embedded controllers faster and much more thoroughly than through the use of other types of testing. The software product to be defined in this project would enable the cost of developing HIL simulations to be reduced significantly, thus encouraging ever greater use of HIL simulations in testing embedded controllers. This would be highly beneficial to the developers of both commercial and military products.
