Ramco OPTIMA

Offers advanced features to drive innovation

The primary task of any optimization package is to manipulate the process variables in order to achieve high efficiency, consistent quality and improved performance in a cost-effective manner. The features built into Ramco OPTIMA helps to achieve these goals. OPTIMA with its open architecture can be interfaced with a wide variety of Process Control and Automation systems. The information acquired is processed through a set of knowledge building blocks called Strategies (Boolean Logic), Fuzzy blocks (Fuzzy Logic) and Control Graphs (Lookup tables). The system then recommends the appropriate values in the process to ensure smooth and stable operations, improved productivity and minimized costs of production. Key features and functionalities offered by Ramco OPTIMA include:

Knowledge Base Building

The process knowledge base comprises computational blocks that can be independently built as required by a process. The knowledge base is typically made up of a number of strategies in a Control engine. Also easy to use editors are provided for the configuration of different type of blocks.

Strategy Editor

Executable statements (rules) are put together in logically related groups to form strategies. Strategies act as a supervisory layer over Fuzzy blocks, Control graphs or other sub-strategies. Control objectives are defined in different strategies and these can be executed collectively or on priorities as required by the process. A library of standard functions (primitives) is provided for building strategy rules.

Fuzzy Logic Editor (Heuristic Modeling)

Process loops where multiple input parameters have an effect on one or more output parameters are best modeled using fuzzy logic. The Fuzzy editor facilitates the building of fuzzy logic blocks for control. Any number of fuzzy blocks can be built and the first step in configuring a fuzzy block is to define the input and output parameters. Then for each parameter, membership sets are configured and users can define the range, shape and number of members for each parameter. Rules for each fuzzy logic block are framed and the entire block is built as per the knowledge elicitation of the process experts. The strategy validates all inputs and outputs of a block before controlling the process. Fuzzy Logic editor provides Ramco OPTIMA the power to perform heuristic modeling.

Control Graphs

Sometimes if a one-to-one relationship is to be depicted and should be used for decision making or control actions, then Control Graphs module is the editor for doing so. Relationships between an independent and a dependent variable can be graphically plotted and can be used inside Ramco OPTIMA for necessary computations or control actions.

Strategy Simulator

Each strategy can be simulated and verified before putting it in operation. On-line process parameters can be inserted in to the strategy. One can see the logical execution of rules, one by one, for prevailing conditions. The user can also enter process values and see the behavior of strategies across the domain.

Fuzzy Block Simulator

This module allows the user to simulate the fuzzy blocks. It helps to fine-tune the blocks, so that only on complete satisfaction, the system is put on-line.

Neural Networks (Data driven modeling)

Neural Networks are useful for the purpose of predicting any process variable’s value, which non-linearly depends on other parameters. Some of the reasons for this were:

  • There is no clearly defined mathematical model available which takes into account all the inputs
  • Neural Nets (NNs) are very good at learning multiple-input relationships.
  • NNs perform very well with noisy data.

Many algorithms are developed to implement neural networks.

The application developed here is used to implement Back propagation with in-house modifications to it for adaptive learning.

Interpolation

In engineering applications, data collected from the field are usually discrete and the physical meanings of the data are not always well known. To estimate the outcomes and, eventually, to have a better understanding of the physical phenomenon, a more analytically controllable function that fits the field data is desirable. The process of finding the coefficients for the fitting function is called curve fitting; the process of estimating the outcomes in between sampled data points is called interpolation; whereas the process of estimating the outcomes beyond the range covered by the existing data is called extrapolation.