Mitsubishi Electric Research Laboratories, Inc.

Mitsubishi Electric Corporation

Ankush Chakrabarty

Stefano Di Cairano

Yebin Wang

Claus Danielson

Karl Berntorp

Rien Quirynen

Scott Bortoff

Mouhacine Benosman

Christopher Laughman

Devesh Jha

Saleh Nabi

Philip Orlik

Jianlin Guo

William T Vetterling

Jonathan Le Roux

Jay E Thornton

Uros Kalabic

Diego Romeres

Abraham Goldsmith

Gordon P Wichern

Marcel Menner

Toshisada Mariyama

Heejin Ahn

Zhong-Qiu Wang

Weinan Gao

Mohit Srinivasan

Yehan Ma

Sanjana VijayShankar

Nobuyuki Yoshikawa

Raphael Chinchilla

William Vetterling

Stefano Dl Cairano

The present disclosure provides a feedback controller and method for controlling an operation of a device at different control steps. The feedback controller comprises at least one processor, and the memory having instructions stored thereon that, when executed by the at least one processor, causes the feedback controller, for a control step, to collect a measurement indicative of a state of the device at the control step, and execute, recursively until a termination condition is met, a probabilistic solver parameterized on a control input to an actuator operating the device to produce a control input for the control step. The feedback controller is further configured to control the actuator operating the device based on the produced control input.

System and method for controlling an operation of a device

A method for controlling a system by a controller comprises accepting a current state of the system and selecting, using a trained function of the current state, a solver from a set of solvers. The method further comprises solving an optimal control optimization problem using the selected solver to produce a current control input, such that for at least some different control steps, the predictive controller solves a formulation of the optimal control optimization problem with different solvers having different accuracies, requiring different computational resources, or both and submitting the current control input to the system thereby changing the current state of the system.

Apparatuses systems and methods for optimization-based control of physical systems

A controller is provided for operating a system under admissible states. The controller includes an interface configured to connect the system storing a set of measured system states, a set of reference inputs and a set of system parameters in a storage arranged inside or outside the system, a memory storing measured system states, admissible reference inputs and admissible parameter sets and computer-executable programs including a parameter estimator and an adaptive reference governor (ARG), a processor, in connection with the memory. The processor is configured to perform the ARG and the parameter estimator. The parameter estimator extracts a pair of a reference input and the system state and compute a system parameter estimate based on the reference input and system state. The ARG is configured to update the reference input and compute a parameter-robust constraint admissible set based on the updated reference input and the system states, wherein the ARG generates and transmits a reference input to the system based on the parameter-robust constraint admissible set.

System and method for data-driven reference generation

A controller is provided for generating a policy controlling a system by learning a dynamics of the system. The controller is configured to perform steps of acquiring measurement data from sensors arranged on the system, providing, to the memory, a non-linear system model represented by known part of the dynamics of the system and unknown part of the dynamics of the system, collecting states of the system by measuring the dynamics of the system using the sensors of the system based on a nominal policy and a noise term with respect to the states, estimating a sequence of sets of states of the system and sets of control inputs by collecting data of the system, wherein the data includes a collection of system states, applied control inputs and change in system states, wherein each of the control input is computed by the nominal policy and the additional noise term, learning a polytopic system by use of the collected data of the system for approximating the unknown part of the dynamics of the system using a linear probabilistic regression model, estimating an attractor basin of a terminal controller by sampling initial states in a neighborhood of a terminal state and estimating the attractor basin by supervised learning, and generating a polytopic policy using the estimated polytopic system to drive the system to the attractor basin of the terminal controller from an initial state.

System and method for polytopic policy optimization for robust feedback control during learning

A controller for controlling a motion of at least one device subject to constraints on the motion, is disclosed. The controller comprises a processor and a memory, where the controller inputs parameters of the task including the state of the at least one device to a neural network trained to output an estimated motion trajectory for performing the task. Further, the controller extracts at least some of the integer values of a solution to a mixed-integer optimization problem for planning an execution of the task that results in the estimated motion trajectory. Further, the controller solves the mixed-integer optimization problem for the parameters of the task with corresponding integer values fixed to the extracted integer values to produce an optimized motion trajectory subject to the constraint and changes the state of the at least one device to track the optimized motion trajectory.

Controller for optimizing motion trajectory to control motion of one or more devices

A system for generating a G-code for controlling an operation of a laser-cutting machine to cut parts from a sheet of material, upon receiving cutting data specifying a cutting order of parts and a cutting order of edges of each part, tests the parts for potential distortions and generates a G-code to avoid the potential distortion. For testing a current part, the system detects a potential distortion when the final edge of the current part is adjacent to an edge of a previously cut part scheduled for cutting before the current part according to the cutting order of parts. The system modifies the cutting order to select the modified cutting order for which the final edge is not adjacent to any edge of any previously cut part.

System and method for detecting and correcting laser-cutting distortion

A system and method for detecting anomalous sound are disclosed. The method includes receiving a spectrogram of an audio signal with elements defined by values in a time-frequency domain of the spectrogram. Each of the values corresponds to an element of the spectrogram that is identified by a coordinate in the time-frequency domain. The time-frequency domain of the spectrogram is partitioned into a context region and a target region. The context region and the target region are processed by a neural network using an attentive neural process to recover values of the spectrogram for elements with coordinates in the target region. The recovered values of the elements of the target region are compared with values of elements of the partitioned target region. An anomaly score is determined based on the comparison. The anomaly score is used for performing a control action.

Method and system for detecting anomalous sound

An extremum seeking control system and a method for controlling a system is provided. The method comprises receiving a tracking error, consecutive measurements of a system output, and consecutive measurements of a cost function of the system, and estimating a gradient based on the tracking error, the consecutive measurements of the system output, and the consecutive measurements of the cost function of the system. The method further comprises determining a step-size based on a ratio of a covariance of the gradient to the gradient, and determining a set-point optimizing performance of the system, based on the gradient and the step-size. The method further comprises controlling the system based on the set-point optimizing performance of the system.

Extremum seeking control system and a method for controlling a system

An apparatus for controlling an operation of a system is provided. The apparatus comprises an input interface configured to receive a state trajectory of the system, and a memory configured to store a model of dynamics of the system including a combination of at least one differential equation and a closure model. The apparatus further comprises a processor configured to update the closure model using reinforcement learning (RL) having a value function reducing a difference between a shape of the received state trajectory and a shape of state trajectory estimated using the model with the updated closure model, and determine a control command based on the model with the updated closure model. Further, the apparatus comprises an output interface configured to transmit the control command to an actuator of the system to control the operation of the system.

Apparatus and method for control with data-driven model adaptation

A system a method for controlling an operation of HVAC system is provided. The system comprises at least one processor; and memory having instructions stored thereon that, when executed by the at least one processor, cause the system to receive data indicative of the operation of the HVAC system and optimize a cost function to reduce a difference between a full thermal state of the conditioned environment simulated from the received data based on a physical model of airflow including a partial differential equation (PDE) and a thermal state of the conditioned environment reconstructed by an observer using a reduced order model (ROM) of airflow including an ordinary differential equation (ODE) from values of the full thermal state at a set of locations, wherein optimization of the cost function is a multivariable optimization of a structure of the observer and a structure of the ODE.

Controlling thermal state of conditioned environment based on multivariable optimization

A vehicle dynamics control system receives a feedback state signal including values of a roll rate and a roll angle of the motion of the vehicle and updates parameters of a model of roll dynamics of the vehicle by fitting the received values into the roll dynamics model. The roll dynamics model explains the evolution of the roll rate and the roll angle based on the parameters including a center of gravity (CoG) parameter modeling a location of a CoG of the vehicle, and a spring constant and a damping coefficient modeling suspension dynamics of the vehicle. The system determines a control command for controlling at least one actuator of the vehicle using a motion model including the updated CoG parameter and submits the control command to the vehicle controller to control the motion of the vehicle.

System and method for adaptive control of vehicle dynamics

A controller for controlling a vapor compression system is provided. The controller is configured to control an operation of the VCS with different combinations of setpoints for different actuators of the VCS to estimate a cost of operation of the VCS for each of the different combinations of setpoints, and compute, using a Bayesian optimization of the combinations of setpoints and their corresponding estimated costs of operation, a probabilistic surrogate model, wherein the probabilistic surrogate model defines at least first two order moments of the cost of operation in the probabilistic mapping. The controller is further configured to select an optimal combination of setpoints having the largest likelihood of being a global minimum at the surrogate model according to an acquisition function of the first two order moments of the cost of operation.

Controlling vapor compression system using probabilistic surrogate model

A system for controlling an autonomous vehicle includes a memory configured to store parameters of a g-g plot defining admissible space of values of longitudinal and lateral accelerations. The g-g plot parameters define a mapping between user driving preferences and constrained control of the autonomous vehicle. The g-g plot parameters include a maximum forward acceleration, a maximum backward acceleration, a maximum lateral acceleration and a shape parameter defining profile of curves connecting maximum values of forward, backward, and lateral accelerations. The system accepts a comfort level as a feedback from a passenger of the vehicle, determines a dominant parameter corresponding to the feedback, updates the dominant parameter of the g-g plot based on the comfort level indicated in the feedback, and controls the vehicle to maintain dynamics of the vehicle within the admissible space defined by the parameters of the updated g-g plot.

Control of autonomous vehicles adaptive to user driving preferences

A control system for controlling a set of actuators of a system. The control system comprising a switcher configured to select a subset of setpoints from a set of setpoints that control the corresponding set of actuators. An extremum-seeking controller (ESC) configured to perturb a subset of setpoints at each iteration based on a probabilistic distribution of partial gradients of a cost function relating values of the subset of setpoints to a cost of operation of the system. A stochastic gradient estimator is configured to estimate a full gradient of the cost function and update the estimation of the full gradient based on the probabilistic distribution of the partial gradients generated at each ESC iteration. A feedback controller is configured to drive a state of the subset of actuators of the system towards the corresponding perturbed subset of setpoints.

Extremum seeking control with stochastic gradient estimation

A computing system for generating optimal tracking control (TC) policies for controlling a machine to track a given time-varying reference (GTVR) trajectory. An updated augmented state of the machine is obtained. Stored in memory is the GTVR trajectory, a constraint-admissible invariant set (CAIS) of machine states satisfying machine state constraints and a corresponding TC policy mapping a machine state within the CAIS to a control input satisfying control input rate constraints. A processor jointly controls the computing system to control the operation to drive an augmented state of the machine to zero, and update the CAIS and TC policy. Joint control includes using a sequence of control inputs and a sequence of augmented machine states within CAIS corresponding to the sequence of tracking control inputs. Execute a constrained tracking approximate dynamic programming (TADP) using the received data to update the value function, update the CAIS and the corresponding TC policy.

Growing community of inventors

Cambridge, MA, United States of America

Ankush Chakrabarty