A system and method for utilizing model predictive control for optimal interactions that include receiving environment e data associated with a surrounding environment of an ego agent and dynamic data associated with an operation of the ego agent. The system and method also include inputting the environment data and the dynamic data to variational autoencoders. The system and method additionally include utilizing the model predictive control through functional approximation with the variational autoencoders and decoders to output probabilistic action estimates. The system and method further include outputting an estimated optimal control trajectory based on analysis of the probabilistic action estimates to control at least one system of the ego agent to operate within the surrounding environment of the ego agent.

A method for reducing and/or managing energy-related stress in an electrical system includes processing electrical measurement data from or derived from energy-related signals captured by at least one intelligent electronic device (IED) in the electrical system to identify and track at least one energy-related transient in the electrical system. An impact of the at least one energy-related transient on equipment in the electrical system is quantified, and one or more transient-related alarms are generated in response to the impact of the at least one energy-related transient being near, within or above a predetermined range of the stress tolerance of the equipment. The transient-related alarms are prioritized based in part on at least one of the stress tolerance of the equipment, the stress associated with one or more transient events, and accumulated energy-related stress on the equipment. One or more actions are taken in the electrical system in response to the transient-related alarms to reduce energy-related stress on the equipment in the electrical system.

A control customization system 80 customizes a plant control. A profiler 81 predicts actions of a user depending on situations of the plant or the user. A planner 82 determines an appropriate set of objectives which represent tasks desired by the user, and objective terms representing elements for controlling the plant so as to realize the objectives, and tunes the objective terms based on predictions of the profiler 81.

Described herein are systems and methods for inverse reinforcement learning to leverage the benefits of model-based optimization method and model-free learning method. Embodiments of a framework combining human behavior model with model predictive control are presented. The framework takes advantage of feature identification capability of a neural network to determine the reward function of model predictive control. Furthermore, embodiments of the present approach are implemented to solve the practical autonomous driving longitudinal control problem with simultaneous preference on safe execution and passenger comfort.

The present invention relates to training predictive data-driven model for predicting an industrial time dependent process. A data driven generative model is introduced for modelling and generating complex sequential data comprising multiple modalities, by learning a joint time-dependent representation of the different modalities. The model may be configured to handle any combination of missing modalities, which enables conditional generation based on known modalities, providing a high degree of control over the properties of the generated sequences.

The present invention relates to an MPC-based hierarchical coordinated control method and device for a wind-hydrogen coupling system. The method comprises the following steps: (1) dividing the wind-hydrogen coupling system into upper-layer grid-connected control and lower-layer electrolytic cell control; (2) controlling grid-connected power to track a wind power prediction curve by adopting an MPC control algorithm for upper-layer grid-connected control, and obtaining an electrolytic cell power control quantity for the lower-layer electrolytic cell control at the same time; (3) dividing operation states of electrolytic cell monomers into four operation states of rated power operation, fluctuating power operation, overload power operation and shutdown; and (4) determining the operation states of various electrolytic cell monomers by adopting a time-power double-line rotation control strategy based on the electrolytic cell power control quantity, thus making the electrolytic cell monomers operate in one of the four operating states in turn.

A device, computer program, and computer-implemented method for determining a variable of a technical system. An input variable is determined for a first model for determining the variable at a first temporal resolution. A first time series is provided, at the first temporal resolution, including values which characterize an operating variable of the technical system. A second time series is provided. at a second temporal resolution, including values which characterize the operating variable of the technical system, the first and second temporal resolutions being different. The second time series is mapped using a second model for determining a first prediction for the variable of the technical system at the second temporal resolution on the first prediction. Parameters of a second model are determined, using the second time series, which are mapped on parameters of a third model at the first temporal resolution.

A method and system for managing electric vehicle (EV) distributed energy resource(s) (DER) are disclosed. A DER analytics engine may receive electricity consumption data of a plurality of sites from corresponding electricity meters of the plurality of sites, detect EV charging information based at least in part on the electricity consumption data, obtain EV telematics data of EVs associated with the EV charging information, reconcile the EV charging information and the EV telematics data, and generate, based on the reconciled EV charging information and the EV telematics data, models for at least one of continuous EV load prediction, electrical vehicle supply equipment (EVSE detection), and/or optimization for at least one of aggregated load, load per feeder, or maximum revenue for time-of-use tiers.

A device for controlling an operation of a system performing a task is disclosed. The device submits a sequence of control inputs to the system thereby changing states of the system according to the task and receives a feedback signal. The device determines a current control input for controlling the system based on the feedback signal including a current measurement of a current state of the system by solving a polynomial optimization of a polynomial function with a reformulation derived by introducing additional variables reducing a degree of the polynomial function till a target degree subject to constraints on a structure of the additional variables. The device solves a mixed-integer optimization problem to find an optimal solution among all possible encodings of factorizations of the polynomial function that reduces the degree of the polynomial function till the target degree with a minimum number of additional variables.

Methods, systems, and computer storage media for providing an optimal control configuration for a material processing system are provided. In operation, a material processing engine accesses causal graph input data. Causal graph input data includes input data of a continuous flow process. Based on the causal graph and the input data, a causal graph that aligns with do-calculus manipulations—associated with determining identifiable causal relationships corresponding to input materials of the continuous flow process—is generated. The causal graph is parsed based on the do-calculus manipulations to determine valid conditioning sets associated with estimating a causal impact on an optimization target. Based on the valid conditioning sets, an optimal control configuration comprising optimal control variable values is generated. Generating the optimal control configuration comprising the optimal control variable values associated with the continuous flow process is based on solving a deterministic convex optimization problem and a corresponding stochastic optimization problem.