When the strip head (7) of a metal strip (1) runs out of a roll stand (2a), a lateral position (y) of the strip head (7) is detected by a detection device (8) at at least one location (P) lying between the roll stand (2a) and a device (8) arranged downstream of the roll stand. A strip position controller (10) is designed as a model predictive controller which ascertains a sequence of adjusting commands (u.sub.k) to be output one after the other in a work cycle (T) on the basis of the detected lateral position (y) of the strip head (7), and the sequence is used to adjust a respective roll gap wedge. The number of control commands (u.sub.k) define a prediction horizon (PH) of the strip position controller (10) in connection with the work cycle (T). The strip position controller (10) at least supplies the roll stand (2a) with the control command (u.sub.0) ascertained to be output next.

A powered balancing mobility device that can provide the user the ability to safely navigate expected environments of daily living including the ability to maneuver in confined spaces and to climb curbs, stairs, and other obstacles, and to travel safely and comfortably in vehicles. The mobility device can provide elevated, balanced travel.

A field device, method, and non-transitory computer readable medium provide for cost of lost opportunity to operators in real-time using an advance process control. The field device includes a memory and a processor operably connected to the memory. The processor receives current values and average values for controlled variables and manipulated variables; determines costs of lost opportunity for each of controlled variable variance issues, limit issues, model quality issues, inferential quality issues, and variable model issues based on the current values and the average values of the controlled variables; and stores the costs of lost opportunity for the field device.

A control apparatus includes a prediction unit configured to repeatedly predict a first target value based on prediction information; a transmission/reception unit configured to repeatedly transmit the prediction information to a server and receive a second target value having higher prediction accuracy than the first target value predicted by the server; a management unit configured to update a first error of prediction in the prediction unit based on the second target value and the first target value; and a setting unit configured to set a control target value based on the first target value and the first error. A first time interval in which the prediction unit repeatedly predicts the first target value is shorter than a second time interval in which the transmission/reception unit repeatedly transmits the prediction information to the server.

A machine-learning control system comprising an operating environment and a sensorium informationally coupled the operating environment. The sensorium comprises a set of sensors and a set of motors, both informationally coupled to a homeostatic network capable of achieving ultrastability within the operating environment. The control system builds a generative model of the operating environment by extracting, through sensorimotor feedback, state information relevant to network ultrastability associated with a particular control behavior and a set of environmental parameters identified within the operating environment. A modulating sensorimotor carrier wave signal may optionally be used to increase training speed of the machine-learning control system. The control system is adaptable to a variety of engineering solutions for autonomous control systems and data processing, such as, for example, autonomous vehicles, robotics, calibration, language processing, and computer vision. A homeostatic network debugger and automatic network topology generation algorithms using node-splitting conditions and functions are also described.

Techniques are disclosed that enable model predictive control of a robot based on a latent dynamics model and a reward function. In many implementations, the latent space can be divided into a deterministic portion and stochastic portion, allowing the model to be utilized in generating more likely robot trajectories. Additional or alternative implementations include many reward functions, where each reward function corresponds to a different robot task.

The invention provides a method for controlling a wind turbine, including predicting behaviour of one or more wind turbine components such as a wind turbine tower over a prediction horizon using a wind turbine model that describes dynamics of the one or more wind turbine components or states. The method includes determining behavioural constraints associated with operation of the wind turbine, wherein the behavioural constraints are based on operational parameters of the wind turbine such as operating conditions, e.g. wind speed. The method includes using the predicted behaviour of the one or more wind turbine components in a cost function, and optimising the cost function subject to the determined behavioural constraints to determine at least one control output, such as blade pitch control or generator speed control, for controlling operation of the wind turbine.

A wind turbine control unit comprising a control module configured to control an actuator system by outputting a first control signal, wherein the first control signal includes a current control sample value and a predicted control trajectory; the control unit further comprising an upsampling module configured to receive the first control signal from the control module, and to output a second control signal for controlling the actuator system, the second control signal having a higher frequency that the first control signal. The upsampling module calculates the second control signal in dependence on the current control sample value and the predicted control trajectory. The embodiments provide a more accurately reproduced control signal at a higher frequency that is suitable for onward processing which does not suffer from the problems of aliasing and delay that exist with conventional upsampling techniques.

A computing system receives historical data. The historical data comprises physical actions taken in an experiment in a physical environment. The experiment comprises user-defined stages. The historical data comprises a recorded outcome, according to user-defined performance indicator(s) related to the user-defined stages, for each physical action taken in the experiment. The system generates, by a discrete event simulator, a computing representation of a simulated environment of the physical environment. The simulated environment comprises processing stages. The system obtains simulation data. The simulation data comprises simulated actions taken by the discrete event simulator. The simulation data comprises a predicted outcome, according to user-defined performance indicator(s) related to the processing stages, for each simulated action taken by the discrete event simulator. The system validates accuracy of the discrete event simulator at predicting the recorded outcome in the experiment. The system trains a computing agent according to a sequential decision-making algorithm.

There is provided a method for providing a driving related guidance service by a service server. The method includes receiving advanced driver assistance system (ADAS) data of a vehicle related to a specific driving situation of the vehicle, location data of the vehicle, driving data of the vehicle, and a driving image captured during driving of the vehicle from a vehicle terminal device, generating guidance information related to the specific driving situation of the vehicle by analyzing the received data and the driving image, and providing a driving related guidance service for the vehicle using the generated guidance information.