Neil F. Palumbo

Integrated missile guidance and control: a state dependent Riccati differential equation approach

The need to engage tactical ballistic missile (TBM) threats and high performance anti-ship cruise missiles is dictating the design of enhanced performance missile interceptors that can provide a high probability of kill. It can be argued that current interceptor guidance and control (G&C) designs are suboptimal because each of the G&C components are designed separately before they are made to interact together as a single functional unit. Ultimately, integrated G&C (IGC) design techniques might improve interceptor performance because: 1) the implicit interdependency of the classically separate G&C components could provide a positive synergism that is unrealized in the more conventional designs, and 2) an IGC design is formulated as a single optimization problem thus providing a unified approach to interceptor performance optimization. The prototype IGC system discussed in this paper is designed via an approximate solution to the nonlinear disturbance attenuation problem. Furthermore, the integrated controller has been implemented in a high fidelity six-degree-of-freedom (6DOF) missile simulation that incorporates a fully coupled nonlinear aerodynamics model. A high-performance benchmark missile G&C system has also been designed and incorporated to provide performance comparisons. In addition to a discussion of the solution methodology, 6DOF Monte Carlo simulation results are presented that compare the integrated concept to the benchmark G&C system. The simulation results to date show the IGC paradigm reduces both the mean and I-sigma miss statistics as compared to the benchmark system.

An expert system for control system design

The Control Laws Expert Assistant for Rotorcraft (CLEAR) project, which is being performed at Temple University and the Boeing Helicopters Company, is described. CLEAR will assist flight control engineers to design a specific set of flight control systems. The resulting interactive knowledge-based system will analyze mathematical models representing rotorcraft and will design any compensation required to enable the system to meet specifications. There are two major roles that CLEAR plays in the design of a rotorcraft flight control system. First CLEAR determines what kind of compensation, if any, is needed. It then places the compensator at an appropriate location within the system and designs the required compensator. The authors focus on the work being performed at Temple University to give CLEAR its design capability. The rationale for using a knowledge-based system is presented, and the CLEAR architecture is described.<<ETX>>

Developing an expert system to assist in control system design

This paper addresses the problem of how to develop an expert system that is able to design or to assist in the design of a control system. The discussion includes how to structure the expert system, how to choose the relationship between the symbolic and numerical processor and how to divide the work among the expert system, the computer aided control systems analysis package and the design engineer. A prototype intelligent design associated is presented and an example of a compensator design is given.

Considerations in the development of a knowledge-based control systems design associate

The authors describe an intelligent design associate (IDA), with emphasis on the techniques used in developing the associate and the mechanics of its operation. The expert-system architecture is described, along with some of the problems encountered when coupling symbolic and numerical systems. Also considered is how to transfer information between the expert system and the analysis package, as well as how the expert system is able to direct the actual sequence of analyses performed by the CACSD (computer-aided control systems design) package. It is demonstrated that not only is such a coupled system feasible, but it is also replicatable in that it can be created with off-the-shelf, commercially available software. An example is included to show that the system is capable of designing simple compensators.<<ETX>>

Considerations in the development of an intelligent design associative

The authors present a simple single-input, single-output, linear, time-invariant, analog control system design example, using an expert system that acts as the design associate. It analyzes the present system, requests the computer-aided control system design (CACSD) package to perform specific computational tasks, looks at the results, decides what type of compensator (in any) need be added to the system, and applies the appropriate compensation heuristic. In addition, the expert system asks for the designers approval of its intended action and allows the designer to modify its suggested design plan.<<ETX>>

Control Laws Expert Assistant for Rotorcraft (CLEAR)

Abstract not available.

An expert system architecture for flight control system design of rotorcraft

The authors describe the system architecture and the design methodology used which together form the basis of an expert system developed for the flight control system design of rotorcraft. The expert system is a coupled system, using NEXPERT for symbolic processing and CTRL-C for numerical processing. How these packages are coupled is described as is the use of each in the design process.<<ETX>>

A new approach to output feedback eigenstructure assignment by way of a bilinear transformation

The output feedback eigenstructure assignment problem is examined, and a new approach to computing the output feedback gain matrix is presented. A key advantage to this new approach is that the number of unknowns, which must be found such that a solution is obtained, is significantly reduced. Another advantage is that, since there is no subjective component introduced in the formulation of the design method which may inhibit or bias the results in some way, several, and in some cases all, solutions can be generated. These solutions may then be examined and candidate designs may be selected. It is proposed that, if coupled to an application specific expert system, the presented methodology can greatly reduce design time while providing superior designs. An example is given to illustrate the usefulness of the proposed method.<<ETX>>

Knowledge-based design of rotorcraft flight control systems

A knowledge-based system called the Control Laws Expert Assistant for Rotorcraft (CLEAR) was developed to help flight control engineers design laws that allow linearized mathematical models of rotorcraft to meet performance criteria as specified in MILSPEC MIL-H-8501B. The system designs compensators that, when coupled with linear, time-invariant, single-input, single-output, continuous plants, will meet the set of prescribed performance specifications. A modular, hierarchical, knowledge-based system designed using an expert system shell which interacts dynamically with a computer-aided control system analysis package is described. A design problem is presented for illustrative purposes.<<ETX>>