Optimization Considerations Relative To The Use Of Hybrid Knowledge-based Decision Support Systems In Autonomous Vehicle Guidance

Autonomous system operations require special con- sideration of real-time constraints and measurement noise, often in demanding environments. Hybrid architectures can take advantage of the features of both heuristic and optimized control theory approaches to obtain results not achievable by architectures of either type taken alone. Heuristic rule-based systems can be used to formulate cost functions for specific tasks and decide the efficiency of calculating such functions relative to using heuristic rules. We discuss considerations that led us to develop hybrid systems,for air, land and sea operations, describe the generic architecture we have developed, and present results of simulation studies.