Machine learning-based systems for parameter deviation monitoring and alerting in engines and pump controls

A system comprising:

• a pump;
• an electric motor configured to be operable for driving the pump;
• a variable frequency drive (VFD) configured to be operable for controlling a speed of the electric motor;
• and
• a controller configured to be operable for controlling the variable frequency drive, the controller further configured to learn, via machine learning, normal power consumption and torque levels of the electric motor across a full operational range of the electric motor, whereby the controller is operable for comparing monitored power consumption and torque levels of the electric motor to the learned normal power consumption and torque levels across the full operational range of the electric motor for detecting an issue(s) or problem(s) associated with the monitored power consumption and torque levels at any given RPM of the electric motor that is within the full operational range of the electric motor;
• wherein the controller is configured to generate an alert and/or initiate a shutdown when the comparison of the monitored power consumption and torque levels of the electric motor to the learned normal power consumption and torque levels indicates an issue(s) or problem(s) associated with the monitored power consumption and torque levels of the electric motor.

A system comprising:

• a first pump;
• a second pump;
• an engine configured to be operable for driving the first pump;
• an electric motor configured to be operable for driving the second pump;
• a variable frequency drive (VFD) configured to be operable for controlling a speed of the electric motor;
• a first controller configured to be operable for controlling the engine, the first controller further configured to learn, via machine learning, normal flow level of the pump across a full operational range of the engine, whereby the first controller is operable for comparing a monitored flow level of the first pump to the learned normal flow level across the full operational range of the engine for detecting an issue(s) or problem(s) associated with the monitored flow level of the first pump at any given RPM of the engine that is within the full operational range of the engine, wherein the first controller is configured to generate an alert and/or initiate a shutdown when the comparison of the monitored flow level of the first pump to the learned normal flow level indicates an issue(s) or problem(s) associated with the monitored flow level of the first pump;
• and
• a second controller configured to be operable for controlling the variable frequency drive, the second controller further configured to learn, via machine learning, normal power consumption and torque levels of the electric motor across a full operational range of the electric motor, whereby the second controller is operable for comparing monitored power consumption and torque levels of the electric motor to the learned normal power consumption and torque levels across the full operational range of the electric motor for detecting an issue(s) or problem(s) associated with the monitored power consumption and torque levels at any given RPM of the electric motor that is within the full operational range of the electric motor, wherein the second controller is configured to generate an alert and/or initiate a shutdown when the comparison of the monitored power consumption and torque levels of the electric motor to the learned normal power consumption and torque levels indicates an issue(s) or problem(s) associated with the monitored power consumption and torque levels of the electric motor.

A system comprising:

• a first pump;
• a second pump;
• an engine configured to be operable for driving the first pump;
• an electric motor configured to be operable for driving the second pump;
• a variable frequency drive (VFD) configured to be operable for controlling a speed of the electric motor;
• and
• a controller configured to be operable for controlling the engine and for controlling the variable frequency drive, the controller further configured to:
• learn, via machine learning, normal flow level of the first pump across a full operational range of the engine, whereby the controller is operable for comparing a monitored flow level of the first pump to the learned normal flow level across the full operational range of the engine for detecting an issue(s) or problem(s) associated with the monitored flow level of the first pump at any given RPM of the engine that is within the full operational range of the engine, wherein the controller is configured to generate an alert and/or initiate a shutdown when the comparison of the monitored flow level of the first pump to the learned normal flow level indicates an issue(s) or problem(s) associated with the monitored flow level of the first pump;
• and
• learn, via machine learning, normal power consumption and torque levels of the electric motor across a full operational range of the electric motor, whereby the controller is operable for comparing monitored power consumption and torque levels of the electric motor to the learned normal power consumption and torque levels across the full operational range of the electric motor for detecting an issue(s) or problem(s) associated with the monitored power consumption and torque levels at any given RPM of the electric motor that is within the full operational range of the electric motor, wherein the controller is configured to generate an alert and/or initiate a shutdown when the comparison of the monitored power consumption and torque levels of the electric motor to the learned normal power consumption and torque levels indicates an issue(s) or problem(s) associated with the monitored power consumption and torque levels of the electric motor.

A method comprising:

• machine learning, via a controller, normal flow level of a pump across a full operational range of an engine, the controller is configured to be operable for controlling the engine for driving a pump;
• monitoring, via the controller, a flow level of the pump;
• comparing, via the controller, the monitored flow level to the learned normal flow level to determine whether the monitored flow level is within an acceptable range or error threshold at a specific RPM of the engine;
• and
• generating, via the controller, an alert and/or initiating a shutdown, via the controller, if the monitored flow level is determined to be outside the acceptable range or error threshold at the specific RPM of the engine.

A method comprising:

• machine learning, via a controller, normal power consumption and torque levels of an electric motor across a full operational range of the electric motor, the controller configured to be operable for controlling a variable frequency drive and the electric motor for driving a pump;
• monitoring, via the controller, power consumption and torque of the electric motor;
• comparing, via the controller, the monitored power consumption and torque to the learned normal power consumption and torque levels to determine whether the monitored power consumption and torque levels is outside an acceptable range or error threshold at a specific RPM of the electric motor;
• and
• generating, via the controller, an alert and/or initiating a shutdown, via the controller, if the monitored power consumption and torque is determined to be outside the acceptable range or error threshold at the specific RPM of the engine.