Neural network circuitry for motors

1. An apparatus for driving a motor, the apparatus comprising:

a plurality of neurons of neural network circuitry implemented using one or more processors and configured to generate a cycle value based on a target speed, based on a speed value associated with the motor at a particular time, and based on a current value associated with the motor at the particular time, wherein the plurality of neurons is configured to be trained to generate the cycle value to minimize an error between the cycle value and a training cycle value for each training vector of a plurality of training vectors and wherein the apparatus is configured to have generated the plurality of training vectors; and

motor circuitry configured to control, based on the cycle value, a set of switching elements to drive the motor, wherein the motor comprises a DC brushless or a DC-excited motor.

2. The apparatus of claim 1 , wherein the apparatus is configured to have generated the plurality of training vectors based on at least one of a range of target speed values for the motor or a range of target current values for the motor that have been determined based on a range of torque values.

3. The apparatus of claim 1 , wherein the neural network circuitry is configured to, for each training vector of the plurality of training vectors, have output the training cycle value to the motor circuitry and to have determined a resulting speed value associated with the motor that occurs when the training cycle value has been output to the motor circuitry.

4. An apparatus for driving a motor, the apparatus comprising:

a plurality of neurons of neural network circuitry implemented using one or more processors and configured to generate a cycle value based on a target speed, based on a speed value associated with the motor at a particular time, and based on a current value associated with the motor at the particular time, wherein the plurality of neurons is configured to be trained to generate the cycle value to minimize an error between the cycle value and a training cycle value for each training vector of a plurality of training vectors and wherein the apparatus is configured to have generated the plurality of training vectors; and

motor circuitry configured to:

control, based on the cycle value, a set of switching elements to drive the motor;

measure a back-electromagnetic force voltage at the motor and at the particular time; and

determine the speed value associated with the motor at the particular time based on the back-electromagnetic force voltage.

5. The apparatus of claim 1 , wherein, to generate the cycle value, the neural network is configured to generate the cycle value further based on a previous speed value associated with the motor at a previous time that occurred before the particular time and a previous current value associated with the motor at the previous time.

6. An apparatus for driving a motor, the apparatus comprising:

a plurality of neurons of neural network circuitry implemented using one or more processors and configured to generate a cycle value based on a target speed, based on a speed value associated with the motor at a particular time, and based on a current value associated with the motor at the particular time, wherein the plurality of neurons is configured to be trained to generate the cycle value to minimize an error between the cycle value and a training cycle value for each training vector of a plurality of training vectors and wherein the apparatus is configured to have generated the plurality of training vectors; and

motor circuitry configured to:

control, based on the cycle value, a set of switching elements to drive the motor;

receive a reference speed; and

determine the target speed based on the reference speed, the speed value associated with the motor at the particular time, and one or more previous speed values associated with the motor at one or more previous times that occur before the particular time.

7. The apparatus of claim 1 , wherein the apparatus is configured to:

generate the plurality of training vectors; and

train the plurality of neurons to generate the cycle value to minimize the error between the cycle value and the training cycle value for each training vector of the plurality of training vectors.

8. The apparatus of claim 7 , wherein, to generate the plurality of training vectors, the apparatus is configured to, for each training vector of the plurality of training vectors:

determine a previous speed value associated with the motor at a first previous time;

determine a previous current value associated with the motor at the first previous time;

determine a previous cycle value for the motor at a second previous time that occurs before the first previous time;

output, at a current time that occurs after the previous time, the training cycle value to the motor circuitry; and

determine a resulting speed value associated with the motor that occurs when the training cycle value has been output to the motor circuitry, wherein, to train the plurality of neurons, the neural network circuitry, is configured to train the plurality of neurons to generate, based on the previous speed value, the previous current value, the previous duty cycle, and the resulting speed value, the cycle value to correspond to the training cycle.

9. The apparatus of claim 8 , wherein, to generate the plurality of training vectors, the apparatus is configured to limit speed at the motor to a range of operating speed values for the motor and limit current at the motor to a range of operating current values for the motor.

10. The apparatus of claim 8 , wherein, to generate the plurality of training vectors, the apparatus is configured to, for each training vector of the plurality of training vector:

determine a second previous speed value associated with the motor at the second previous time that occurred before the previous time;

determine a second previous current value associated with the motor at the second previous time; and

determine a second previous cycle value for the motor at a third previous time that occurred before the second previous time, wherein, to train the plurality of neurons, the neural network circuitry, is configured to train the plurality of neurons to generate, based further on the second previous speed value, the second previous current value, and the second previous duty cycle, the cycle value to correspond to the training cycle.

11. The apparatus of claim 7 , wherein a variable mechanical load is applied to the motor when the neural network trains the plurality of neurons.

12. The apparatus of claim 11 , wherein a load motor mechanically coupled to the motor is configured to apply the variable mechanical load to the motor when the neural network trains the plurality of neurons.

13. The apparatus of claim 1 , wherein, to control, based on the cycle value, the set of switching elements to drive the motor, the motor circuitry is configured to:

generate, based on the cycle value, a digital modulated signal; and

drive, based on the digital modulated signal, a set of switching elements to operate in at least a first switching state and a second switching state, wherein, during the first switching state, the set of switching elements electrically couples a first terminal of the motor and a first supply terminal of a supply and electrically couples a second terminal of the motor and a second supply terminal of the supply and wherein, during the second switching state, the set of switching elements electrically couples the second terminal of the motor and the first supply terminal and electrically couples the second terminal of the motor and the first supply terminal.

14. A method for driving a motor, the method comprising:

generating, by a plurality of neurons of neural network circuitry implemented using one or more processors and of an apparatus for driving the motor, a cycle value based on a target speed, based on a speed value associated with the motor at a particular time, and based on a current value associated with the motor at the particular time, wherein the plurality of neurons is configured to be trained to generate the cycle value to minimize an error between the cycle value and a training cycle value for each training vector of a plurality of training vectors and wherein the apparatus is configured to have generated the plurality of training vectors; and

controlling, by motor circuitry and based on the cycle value, a set of switching elements to drive the motor, wherein the motor comprises a DC brushless or a DC-excited motor.

15. The method of claim 14 , wherein the apparatus is configured to have generated the plurality of training vectors based on at least one of a range of target speed values for the motor or a range of target current values for the motor that have been determined based on a range of torque values.

16. The method of claim 14 , wherein the neural network circuitry is configured to, for each training vector of the plurality of training vectors, have output the training cycle value to the motor circuitry and to have determined a resulting speed value associated with the motor that occurs when the training cycle value has been output to the motor circuitry.

17. The method of claim 14 , comprising:

measuring, by speed circuitry, a back-electromagnetic force voltage at the motor and at the particular time; and

determining, by the speed circuitry, the speed value associated with the motor at the particular time based on the back-electromagnetic force voltage.

18. The method of claim 14 , wherein generating the cycle value comprises generating the cycle value further based on a previous speed value associated with the motor at a previous time that occurred before the particular time and a previous current value associated with the motor at the previous time.