A motor control method inputs one or more controlled variables or target values each representing a state of a motor to one or more node layers as an input value, and performs calculation in each of the one or more node layers to output one or more manipulated variables used for control of the motor and control the motor in accordance with the one or more manipulated variables. Each the one or more node layers has a plurality of nodes that execute calculations in parallel. Each of the plurality of nodes multiplies the input value by a coefficient specified for the corresponding node, and performs calculation using a function specified for the corresponding node and designating a multiplied value as an input variable to determine an output value.

This power conversion device includes: a current detection unit for detecting current flowing through a rotary electric machine; a switching pattern determination unit for determining a switching pattern on the basis of the detected current, a current prediction value, a current command value, and a current harmonic command value; and a power conversion unit for outputting AC power to the rotary electric machine in accordance with the switching pattern, wherein the switching pattern determination unit determines the switching pattern so that the current value follows the current command value and the current harmonic becomes equal to or smaller than a limit value.

A motor control method inputs one or more controlled variables or target values each representing a state of a motor to one or more node layers as an input value, and performs calculation in each of the one or more node layers to output one or more manipulated variables used for control of the motor and control the motor in accordance with the one or more manipulated variables. Each the one or more node layers has a plurality of nodes that execute calculations in parallel. Each of the plurality of nodes multiplies the input value by a coefficient specified for the corresponding node, and performs calculation using a function specified for the corresponding node and designating a multiplied value as an input variable to determine an output value.

Provided is a motor control device capable of improving efficiency in real time by a neural network structure that directly derives, in a learning manner, an output signal providing optimal efficiency. A motor control device 1 is adapted to control a motor 6, and includes a neural network compensator 11 that receives input signals and repeats learning based on forward propagation and backpropagation thereby to derive an output signal providing optimal efficiency. Input signals are a motor current, a motor parameter and torque, and the like, and output signals are a current command value and a current phase command value. The motor 6 is controlled on the basis of an output signal derived by the neural network compensator 11.

The disclosure discloses a motor driving system converter fault diagnosis method based on adaptive sparse filtering, and belongs to the field of driving system fault diagnosis. The disclosure applies an unsupervised learning algorithm to an application scene of converter fault diagnosis. Effective features are automatically extracted from original data, and the problem of manual feature design based on expert knowledge is solved. Meanwhile, in consideration of current fundamental period change caused by different rotation speed working conditions, rotation speed feedback is introduced, secondary sampling is carried out on current sampled at a constant frequency, it is ensured that the length of a signal input into the deep sparse filtering network is one fundamental wave period, redundant information is better removed from original data, the calculation burden is relieved, and the accuracy and rapidity of the diagnosis algorithm are improved to a certain extent.

A method for detecting a structural fault of an electric motor. The method includes: (i) acquiring a measurement signal of a physical parameter representative of the rotation of the electric motor, (ii) obtaining the high frequency component of the measurement signal, resulting in a corrected signal, (iii) applying a set of band-pass filters to the corrected signal resulting in a set of filtered corrected signals, (iv) determining a stator frequency and a rotor frequency, (v) computing a frequency signature vector, (vi) computing a temporal symptom vector from the frequency signature vector and the set of filtered corrected signals, and (vii) detecting a structural fault of the electric motor from the determined temporal symptom vector and from a classifier model.

A control system for controlling a power generator of a power generation system executes a control policy to map an input-and-output sequence to a current value of the excitation voltage, submits the current value of the excitation voltage to the power generator, accepts a current value of the rotor angle caused by actuating the power generator according to the current value of the excitation voltage, and updates the input-and-output sequence with the corresponding current values of the rotor angle and the excitation voltage. The input-and-output sequence of values of the operation of the power generator includes a sequence of multiple values of the rotor angle of the power generator and a corresponding sequence of multiple values of excitation voltage to the power generator causing the values of the rotor angle. The control policy maps the input-and-output sequence to a current control input defining the current value of the excitation voltage.

An apparatus for driving a motor includes motor circuitry and neural network circuitry. The motor circuitry is configured to generate, based on an error compensated rotor angle and current at a plurality of phases of the motor, a d-axis instant current value and generate a d-axis instant voltage value based on the d-axis instant current value. The motor circuitry is further configured to generate voltage at the plurality of phases based on the d-axis instant voltage value. The neural network circuitry is configured to generate a rotor angle offset based on an instant rotor speed at the motor. The neural network circuitry has been trained to generate the rotor angle offset to minimize the d-axis instant voltage value for each of a plurality of rotor speeds at the motor. The error compensated rotor angle is based on the rotor angle offset.

A machine learning apparatus for learning a correction parameter used in correction of a command value that controls a motor in a motor drive system including a plurality of kinds of correction functions includes: a state observation unit that observes, as a state variable, each of a feature calculated on the basis of drive data and the kind of any of the correction functions of the motor drive system and the correction parameter; and a learning unit that learns the correction parameter for each of the correction functions according to a training data set created on the basis of the state variable.

A machine learning apparatus for learning a correction parameter used in correction of a command value that controls a motor in a motor drive system including a plurality of kinds of correction functions includes: a state observation unit that observes, as a state variable, each of a feature calculated on the basis of drive data and the kind of any of the correction functions of the motor drive system and the correction parameter; and a learning unit that learns the correction parameter for each of the correction functions according to a training data set created on the basis of the state variable.