In this paper, we present the application of the Linear Quadratic Gaussian (LQG) controller to improve the performance of weighcell based checkweighing system. The LQG controller was designed and its characteristics presented. The method was also compared with the results of previous work.

The plant that should be controlled is a weighcell based dynamic weighing system. The weighcell is a controllable weighing device that operates according to the principle of electromagnetic force compensation. The two main aims for improvement are: (1) to increase the speed of weighing and (2) to achieve good measurement accuracy. These goals are contradictory and can be completed through an integrated control and filtering approach. The Linear Quadratic Gaussian (LQG) is an optimal design method that combines the design of both control and filtering in an integrated manner. Therefore, this paper discusses a possibility of using LQG method for enhancement of the performance of weighcell based dynamic weighing systems. Furthermore, the paper includes mathematical models of the weighcell and forcing functions. These models are used to examine the suitability of the proposed LQG design method

The objective of this note is to present a solution to the decentralized estimation and control problem for linear discrete-time varying systems, composed of overlapping subsystems. The solution is based on the expansion-contraction framework of the inclusion principle. It is shown how decentralized estimation and control laws can be independently computed for the expanded system, and then contracted for implementation in the original system to satisfy the overlapping information structure constraint.

In this report, a ''piece-by-piece'' scheme is proposed for a hierarchical decomposition and simulation of sparse large-scale systems. Advantage is taken of the special lower-block triangular structure of the system matrices resulting from the application of a graph-theoretic decomposition algorithm. We consider any number of subsystems in the hierarchy, and build the overall system simulator as a union of local, subsystem simulators resulting in the faster overall system simulation. The proposed scheme is both off-line and on-line computationally efficient, which is especially attractive in the case of large-sparse systems. Further computational savings can be achieved by implementing the hierarchical simulation algorithm using multiprocessor computer systems. 23 refs., 10 figs., 2 tabs.

In this report, a ''piece-by-piece'' scheme is proposed for a hierarchical LQG design of estimators and controllers for sparse large-scale systems. Advantage is taken of the special lower-block triangular structure of the system matrices resulting from the application of a graph-theoretic decomposition algorithm. We consider any number of subsystems in the hierarchy, and build the overall system estimator and controller as a union of locally optimal, subsystem estimators and controllers. Due to the hierarchical structure of the decomposed system, the overall closed-loop system is stable and suboptimal. 28 refs., 20 figs.

The objective of this paper is to present a solution to the decentralized estimation and control problem for linear discrete time-varying systems, composed of overlapping sub-systems. The solution is based on the expansion-contraction framework of the Inclusion Principle. It is shown how decentralized estimation and control laws can be independently computed for the expanded system, and then contracted for implementation in the original system to satisfy the overlapping information structure constraint.