Computational architecture and method for a time-varying control system

Abstract

Automatic control of an enterprise. A control vector is generated corresponding to a target allocation of resources between the plurality of products to be produced by the enterprise. An observation vector represents actual retained earnings attributed to each of the plurality of products during a first time period, the actual retained earnings being based on data representing the actual revenue and on data representing the actual cost. A dynamic matrix, a control matrix, a cost matrix, and an observation matrix are produced. Each matrix represents a corresponding dynamic relationship between a sets of different control parameters or measured or predicted values, and is revised at each subsequent time period to produce predicted values for each corresponding matrix.According to various embodiments, the revision can have two components: (a) vectorization; and (b) forecasting.

Description

PRIOR APPLICATION[0001]This application is a continuation of and claims priority to U.S. patent application Ser. No. 14 / 829,311, filed on Aug. 18, 2015, which claims priority to U.S. Provisional Application No. 62 / 038,600 filed on Aug. 18, 2014, the benefit of priority of each of which is claimed hereby, and each of which are incorporated by reference herein in its entirety.FIELD OF THE INVENTION[0002]The invention relates generally to an adaptive time varying control system, and more particularly to the use of computational models and machine-executed simulations to predict the operational stability of an enterprise and its production performance. Embodiments of the invention relate to improving the working computational models that represent the operational performance of an enterprise and implementing those models in specialized computer systems.BACKGROUND OF THE INVENTION[0003]Enterprises, such as producers of goods or services, systematically allocate resources between a variet...

AI Technical Summary

Benefits of technology

[0016]According to certain embodiments, modeling of complex, non-stationary systems is avoided by virtue of the revising of the matrices to produce predicted mappings between the various parameters. In one aspect of the invention, the utilization of the cost matrix takes exogenous factors into account in a computationally-efficient and practical manner. For example, for economic applications the cost matrix will map the cost vector depending on demand and current market state to the system's state vector. Further computational efficiencies are gained in the obtaining of the control vector (also called control law) affecting the system's performance using numerical techniques based on the measured (current time period) and estimated (next time period) state and cost vectors and on updated and estimated matrix values, thereby avoiding dependency on any the myriad of difficult-to-model parameters.

Problems solved by technology

Often, the marginal cost of production of each product or service varies, as does the selling price for that product.

Finding this optimal allocation has been an immensely difficult challenge for enterprise operators.

However, in situations where the control system is using a mathematical enterprise model and variables affecting the optimal allocation of resources are themselves functions of time, known control systems tend to be unable to provide complete guidance on the proper allocation of resources.

Another challenge with optimizing, or otherwise controlling, the allocation of resources within an enterprise is achieving usable, accurate modeling of complex dynamical systems or non-stationary processes.

When computational tools are applied in areas such as economics, biology, medicine, and the like, where the systems are subject to myriad time-varying influences from numerous, often inter-related, parameters, conventional control system techniques, which involve modeling of the system, tend to fall short for lack of modeling accuracy.

Images