October 24, 2019 – 10am (Detroit) | 4pm (Stuttgart) | 7:30pm (New Delhi)
Register for this webinar
If you can’t attend the session live please register anyway and we’ll send you a link to the slides and a video of the session when it’s finished.
As vehicle complexity increases, the number of product lines grows and performance must be assured across many real-world scenarios, the cost and time required to calibrate conventional control systems is increasing exponentially.
In this 60-minute webinar, Asif Farooq, Senior Project Engineer, Control & Calibration and Peter Martin, Technical Specialist, Control & Calibration of Ricardo, introduce Model Predictive Control (MPC). This approach embeds a model of the system being controlled into an electronic control unit (ECU), which is used to predict system response. An online optimisation uses this model to find the best choice of control signal at each update, while operating within physical constraints found in any practical system.
Conventional controllers often require extensive calibration to achieve good performance within these limits. By contrast, MPC optimisation and constraint handling typically results in a shorter calibration process, with reduced development time and cost to achieve good performance. Ricardo’s modelling approach helps to accelerate the development process by reducing complex plant models to simplified control-oriented models that can be deployed on an ECU. Control-oriented models are developed using system identification, physical equations or a combination of the two.
MPC can be applied to many systems within a vehicle, including engine control, thermal system control, cruise control, energy management and more. As vehicles become more connected, predictive controllers can also exploit available data on future operating conditions to further improve performance.
Additional insight is also shared on Ricardo’s broad experience with Model Predictive Control, describing where it is best to apply the technique and drawing on examples from a number of vehicle applications.