MPC700Model Predictive Control Course

Build, commission, and maintain multivariable MPC controllers using closed-loop time-domain identification - 2 days in the classroom or 18 hours online, from step testing and model building through commissioning and long-term maintenance.

Request Course Info What you'll learn

COLUMBO closed-loop MPC identification softwareCompletion certificate on course finish

Overview

Industrial Model Predictive Control

The MPC700 course trains process control engineers, DCS technicians, and supervisors to build, commission, and maintain multivariable model predictive control (MPC) systems, using PiControl's COLUMBO software to identify and refresh dynamic models directly from closed-loop, time-domain plant data. MPC700 runs as 2 days of classroom training or 18 hours online, with training slides and MPC software from the first session.

The course starts from the history and case for MPC, where MPC outperforms single-loop and advanced regulatory control, and where a simpler methodology is the better fit. It then covers conducting step tests and identifying MPC models, designing and building the MPC controller, and startup and commissioning, before moving into MPC maintenance - modifying and improving models after years of operation or after significant process changes, including automated step testing, PRBS, and other modern identification techniques. Attendees finish able to design, maintain, and troubleshoot MPC controllers, and to use modern closed-loop dynamics identification with COLUMBO to improve MPC models without new open-loop step-testing campaigns.

Curriculum

What You Learn in the MPC700 Model Predictive Control Course

MPC700 moves from the fundamentals of MPC through model identification, controller design, and commissioning, and into long-term maintenance - with hands-on work on COLUMBO throughout. After completing the course, attendees can design, build, and troubleshoot MPC controllers, decide when MPC is the right tool versus APC or single-loop control, and use modern closed-loop, time-domain identification to keep MPC models accurate for years after commissioning, not just at project handoff.

The course covers the following topics.

Module 01 · 5 topics

MPC and APC fundamentals

APC and MPC - history, need, and where each excels
Components of an MPC system
How MPC works
When to use, and when not to use, MPC
Choosing MPC or APC for a given process

Module 02 · 5 topics

Dynamic models and step testing

Dynamic models and step response coefficients
Conducting step tests in the plant
Slave PID control optimization
The MPC control matrix
MPC tuning parameters

Module 03 · 4 topics

Building the controller with COLUMBO

LP costs and optimization
Let's build a new MPC system - hands-on design labs
Closed-loop, time-domain model identification labs on COLUMBO
Verifying MPC predictions against plant data before startup

Module 04 · 4 topics

Commissioning and MPC maintenance

MPC commissioning and good practices
MPC monitoring and troubleshooting
Automated step testing, PRBS, and other modern identification techniques
Modifying and improving MPC models after years of operation or major process changes

Once these topics are complete, engineers choose the delivery format that fits their schedule and plant access.

Who it's for

Who the Course Is For

MPC700 is built for process control engineers, DCS technicians, and supervisors who design, run, or oversee multivariable MPC controllers. Attendees should have knowledge of primary process control and PID loops, and preferably a few months of plant experience, especially on a DCS, because the course moves quickly from MPC fundamentals into applied model building and commissioning.

Process control engineers

Learn to conduct step tests, identify dynamic models, design the MPC control matrix, and build a new MPC system from the ground up, then keep it accurate with closed-loop identification long after commissioning.

DCS technicians

Gain the skills to observe plant trends, recognize a degrading MPC controller, run step tests correctly, and troubleshoot the controller alongside operations and control engineers.

Supervisors

Learn when MPC is the right choice over APC or single-loop control, how to quantify the business case for a project, and what good commissioning and maintenance practice looks like before approving one.

Bring your whole team

MPC700 can be tailored for in-plant team training with plant-specific cases, using your own plant's data, models, and DCS platform for an applied learning experience. Group participation works well when a team is responsible for MPC commissioning, DCS migrations, or advanced control upgrades. Onsite corporate training is available on request so a team trains together on its own controller.

Schedule On-Site Training

Hands-on labs

Hands-On With COLUMBO

Hands-on practice separates a working MPC course from a lecture on theory. MPC700 trains engineers on COLUMBO, PiControl's closed-loop universal multivariable optimizer, which identifies and refreshes MPC dynamic models directly from closed-loop, time-domain historian data - including data with slave PID controllers in auto or cascade mode. That removes the need for a fresh open-loop step-testing campaign every time a model needs refitting, so attendees learn to build and maintain MPC models the way they will on the job.

The course also covers conventional step testing, the MPC control matrix, LP cost optimization, and commissioning practice, so attendees can build a new MPC system as well as maintain and improve an existing one. COLUMBO is vendor-independent, so the same closed-loop identification skills apply whether the controller in the plant is Aspen DMC, Honeywell RMPCT, Emerson PredictPro, Yokogawa PACE/SMOC, or Foxboro Aveva Connoisseur. Ongoing MPC health work after the course is covered in PiControl's MPC maintenance and improvement service.

COLUMBO · closed-loop model identification used throughout MPC700

Certification & materials

Certification and Course Materials

MPC700 attendees receive a PiControl MPC700 Completion Certificate, which documents completion of the course for professional development records or a resume or LinkedIn profile. Attendees also receive documentation templates, commissioning checklists, and model validation sheets to reuse when building, commissioning, or maintaining MPC systems at their own plant. Engineers who want to keep building APC and MPC skills can continue toward broader MPC certification.

Course materials include training slides and MPC software, temporary software access, and follow-up guidance after the course. PiControl also offers APC200 for engineers who want a broader foundation in advanced process control alongside their MPC skills, and remote MPC maintenance and improvement consulting for plants that need a controller refit rather than, or in addition to, training.

Completion certificate
Documentation templates
Commissioning checklists
Model validation sheets
Temporary software access
Follow-up guidance

Request Course Info

PiControl Solutions

Certificate of Completion

MPC700

Industrial Model Predictive Control

Instructor

Date

FAQ

MPC700 Frequently Asked Questions

Short answers to the questions engineers ask most before enrolling in the MPC700 Model Predictive Control course.

Request Course Info

MPC700 provides in-depth training on Model Predictive Control (MPC), covering its design, implementation, commissioning, troubleshooting, and maintenance. It teaches how to use MPC software, including COLUMBO, to build multivariable controllers and keep them optimized in industrial settings.

MPC700 runs as 2 days of classroom training or 18 hours online. The classroom format is instructor-led, and the online format is self-paced so engineers can complete it around plant duties.

MPC700 is designed for process control engineers, DCS technicians, and supervisors with prior knowledge of primary process control and PID loops, and ideally some plant experience, especially with DCS systems.

Attendees should have a basic understanding of process control, including PID control, and preferably a few months of plant experience, especially with DCS systems. MPC700 starts from the fundamentals of MPC before moving into applied model building and commissioning.

MPC700 uses COLUMBO, PiControl's MPC software, plus training slides. Attendees identify models, build the control matrix, and commission a controller on the real-time software rather than working only from theory.

MPC is an advanced, multivariable control strategy that uses dynamic process models to predict future plant behavior and make coordinated, real-time adjustments across manipulated variables while respecting hard constraints. It is well suited to complex, interacting processes.

MPC700 covers the strengths and weaknesses of both MPC and APC directly, so attendees can weigh loop interactions, constraint handling, and project cost to choose the most suitable strategy for a given process.

Yes. MPC systems need regular maintenance to stay effective, especially after process changes. MPC700 teaches how to modify and improve MPC models over time using closed-loop, time-domain identification with COLUMBO, including automated step testing and PRBS techniques, so a controller keeps performing well after the original project team has moved on.

Yes. MPC700 walks through how to quantify the business impact of MPC, such as improved yields, reduced energy use, and less downtime, making it easier to present a business case to management before committing to a project.

Attendees receive a PiControl MPC700 Completion Certificate, along with documentation templates, commissioning checklists, and model validation sheets to reuse at their own plant.

Yes. PiControl offers custom training sessions that incorporate a plant's own data, models, and DCS platform for an applied learning experience, and onsite corporate training is available on request so a team trains together on its own controller.

Get started

Get Started With MPC700

Request course info for MPC700 to give your engineering team a practical, COLUMBO-based Model Predictive Control course with a completion certificate. Online, 2-day classroom, and onsite formats are available, so teams in any location or time zone can start.