Adding Value by Optimizing the Process
Expert Optimizer adds value to the operational performance by
- Reducing variability and,
- Bringing your process closer to its operational limits.
cpmPlus Expert Optimizer provides unmatched performance due to
Use Process Knowledge
- Managing the critical control limits
- Consistently implementing the best strategy
- Executing frequent small changes, as opposed to large step changes
- Immediately recognizing abnormal conditions and acting accordingly
cpmPlus Expert Optimizer helps the plant management to reach its profitability and sustainability goals. Expert Optimizer provides solutions for the process industries including:
- Oil & gas facilities
- Cement grinding plants
- Food and beverage facilities
- Pulp and paper mills
- Industrial power plants
ABB has proven process optimization strategies that are tuned to match the characteristics and needs of the customer site.
In addition to their own process knowledge, customers can tap the vast experience of ABB’s process engineers to develop the strategy best suited to optimize their dynamic process. ABB offers comprehensive training and application support for Expert Optimizer end users.
A comprehensive approach to a challenging problem
The cpmPlus Expert Optimizer toolkit provides a comprehensive variety of advanced control techniques for appropriate strategy development. Coupled to the graphical engineering environment, this ensures fast development and implementation and simplified long term maintenance of a system. The cpmPlus Expert Optimizer toolkit is used to build and display the control strategy required to achieve the process and business objectives.
Expert Optimizer consists of variable gain, multivariable Fuzzy Logic rules, Neural Networks and Model Predictive Control in linear and nonlinear fashions. Process engineers that use cpmPlus Expert Optimizer have successfully achieved more than 4 million hours run time in closed loop control.
Fuzzy Logic / Neural Networks
Fuzzy logic incorporates human knowledge to make and implement effective decisions in a process, while neuro-fuzzy networks are used to learn relationships between key process variables. The integration of these complementary control techniques allows the engineering of powerful robust solutions, which provide substantial financial benefits to the factory for extended periods of time. An understanding of the process is critical to a successful implementation, and ABB’s extensive industrial and project expertise will complement the user’s process knowledge.
Model Predictive Control (MPC)
One of the leading technologies in Expert Optimizer applications is Model Predictive Control (MPC). MPC is based on the predictive capabilities of a mathematical model that allows a sequence of future optimal control actions to be derived. Initial target values are applied to the process, and when new information becomes available, a new sequence is determined. Each sequence is computed by means of an optimization procedure, which follows two goals:
- Optimize the performance
- Protect the system from constraint violations
MPC technology is a widely used industrial technique for advanced multivariable control. For processes featuring strong interaction among different units, MPC offers substantial performance improvements compared with traditional single-input single-output control strategies. Additionally, MPC provides the capability to handle constraints, for both manipulated inputs and process variables. The future process behavior is predicted and compared to process operating objectives and constraints within an optimization engine that computes appropriate future control actions.