Apparatuses, systems, and methods of physical-model based building automation using in-situ regression to optimize control systems are presented. A simulation engine is configured to simulate a behavior or a controlled system using a physical model for the controlled system. A data stream comprises data from a controlled system. A training loop is configured to compare an output of a simulation engine to a data stream using a heuristic so that a physical model is regressed in a manner that the output of the simulation engine approaches the data stream.

An occupancy tracking device configured to establish a network connection with an access point and to capture wireless signal distortion information for the network connection. The device is further configured to generate statistical metadata for the wireless signal distortion information. The device is further configured to input the wireless signal distortion information and the statistical metadata for the wireless signal distortion information into a machine learning model. The machine learning model is configured to determine a predicted occupancy level based on the wireless signal distortion information and the statistical metadata for the wireless signal distortion information. The predicted occupancy level indicates a number of people that are present within with the space. The device is further configured to obtain the predicted occupancy level from the machine learning model and to control a Heating, Ventilation, and Air Conditioning (HVAC) system based on the predicted occupancy level.

Methods and systems for controlling a system such as an engine having an airflow system. A model predictive control calculation is configured in an off-line mode, having a linear part and a non-linear part. In an on-line mode, the linear part of the MPC and/or a Hessian matrix used with the MPC is modified responsive to special modes or other operating changes or conditions. The online mode is configured to respond to changing modes or conditions without requiring recalculation of the MPC. Certain changes of conditions and modes are used to modify feedforward, while others modify responsiveness.

The present disclosure is directed to systems and methods for economically optimal control of an electrical system. Some embodiments employ generalized multivariable constrained continuous optimization techniques to determine an optimal control sequence over a future time domain in the presence of any number of costs, savings opportunities (value streams), and constraints. Some embodiments also include control methods that enable infrequent recalculation of the optimal setpoints. Some embodiments may include a battery degradation model that, working in conjunction with the economic optimizer, enables the most economical use of any type of battery. Some embodiments include techniques for load and generation learning and prediction. Some embodiments include consideration of external data, such as weather.

A method for controlling power consumption includes estimating power flexibility for one or more electrical systems in a building, presenting one or more power flexibility options to a user in the building, and communicating power set points to the one or more electrical systems based on a power flexibility option selected by the user. A system for predicting energy consumption of a building is also disclosed.

A predictive controller controls a system under uncertainty subject to constraints on state and control variables of the system. At each control step, the predictive controller solves an inequality constrained nonlinear dynamic optimization problem including probabilistic chance constraints representing the uncertainty to produce a control command, and controls an operation of the system using the control command. The predictive controller solves the dynamic optimization problem based on a two-level optimization that alternates, until a termination condition is met, propagation of covariance matrices of the probabilistic chance constraints within the prediction horizon for fixed values of the state and control variables with optimization of the state and control variables within the prediction horizon for fixed values of the covariance matrices.

Active learning for operating a physical system. The method includes: providing a data set that comprises data points each comprising an input for operating the physical system, and a first and second observation of the physical system; training a multi-output Gaussian process for predicting the first observation for a given input with the data set; training a Gaussian process for predicting the second observation for a given input with the data set; determining with the data set an input for operating the physical system; determining the first and second observations that result from operating the physical system with the determined input; and adding a data point to the data set that comprises the determined input and the determined first and second observations.

There is provided a technique of managing an electrical power grid. The technique comprises, by a computer: processing timestamped data informative of weather conditions and of individual grid power consumption by a plurality of consumers to identify dual consumers connected to alternative power sources with power generating dependable on the weather conditions; for the dual consumers, forecasting alternative power production by respective connected alternative power sources; and using the provided forecast to enable management action(s) with regard to power production in the electrical power grid (e.g. issuing command(s) related to charging/discharging one or more batteries connected to the grid, controlling thermostat set-point change in a set of points connected to the grid, etc.). Forecasting alternative power production can be provided using a trained Forecasting Machine Learning Model trained to forecast the alternative power production in accordance with a forecast of the one or more weather conditions in the geographical area.

A computer-implemented method, computer program product, and computer system for automated model predictive control. The computer system trains multiple step look-ahead regression models, using historical states and historical actions for a to-be-optimized system, for each timestep of a past time horizon. Regression models may be either linear or nonlinear in order to capture process dynamics and nonlinearity. The computer system generates optimization constraints for each timestep of a future time horizon. The computer system generates optimization variables, based on the multiple step look-ahead regression models, for each timestep of the future time horizon. The computer system constructs a mixed integer linear programming based optimization model that includes an objective function, the optimization constraints, and the optimization variables. Nonlinear regression models are converted into piecewise linear approximation functions. The computer system solves the optimization model to produce actions for the to-be-optimized system, over the future time horizon, and recommend commitment-look-ahead actions.

The present disclosure is directed to machine learning of electrical power system behavior, and related systems, apparatuses, and methods. A controller of an electrical power system includes a data storage device configured to store model data indicating a model load power consumed by loads of the electrical power system. The controller also includes a processor configured to determine current data including current load power consumed by the loads, modify the model data by aggregating the model data with the current data, and determine a set of control values for a set of control variables to effectuate a change to operation of the electrical power system based, at least in part, on the model data.