Systems and methods for operating an uninterruptible power supply (UPS)

1. A system for operating uninterruptible power supplies (UPSs), the system comprising:

a plurality of capacitor banks;

a UPS control circuit coupled to the plurality of capacitor banks, the UPS control circuit configured to:

cause an output of a clean electrical power, wherein the clean electrical power corresponds to an electrical power having voltage fluctuations smoothed out;

sample the clean electrical power at a predetermined sample rate;

calculate control algorithm outputs based on the sampled clean electrical power; and

output control algorithm results corresponding to the calculated control algorithm outputs to a central control circuit, wherein the central control circuit is remote and distinct from the UPS control circuit; and

the central control circuit executing a trained machine learning model configured to:

detect a degradation of the plurality of capacitor banks beyond a health threshold based on the control algorithm results such that the trained machine learning model detects the degradation as a function of at least a number of events when a reverse voltage is applied on the plurality of capacitor banks and a number of events when an excessive voltage is applied on the plurality of capacitor banks; and

transmit a service alert in response to the detection of the degradation of the plurality of capacitor banks.

2. The system of claim 1 , wherein the detection of the degradation is further a function of charge and discharge cycles of the plurality of capacitor banks, charge and discharge durations of the plurality of capacitor banks, capacitor operating voltage, a ripple current value, an internal control coefficient, capacitance value, ambient temperature, capacitor temperature, and an internal coefficient.

3. The system of claim 1 , wherein the clean electrical power comprises a three-phase power.

4. The system of claim 1 , wherein the predetermined sample rate comprises a switching frequency greater than 5 kilohertz (kHz).

5. The system of claim 1 , wherein each UPS unit of a plurality of UPS units comprises the plurality of capacitor banks and the UPS control circuit, and wherein the central control circuit is distinct and remote from the plurality of UPS units and further configured to:

receive each of a plurality of measurements, coefficients, and control data from each UPS unit at a particular time; and

cause a database to store the plurality of measurements, coefficients, and control data.

6. The system of claim 1 , wherein the control algorithm results comprise capacitor operating voltages associated with each UPS unit of a plurality of UPS units at half load.

7. The system of claim 1 , wherein the control algorithm results comprise ripple current values associated with each UPS unit of a plurality of UPS units measured at an output filter of the UPS unit prior to a receipt of an alternating current (AC) output at a load.

8. The system of claim 1 , further comprising a display unit configured to display the service alert prompting a user to provide a maintenance service to the plurality of capacitor banks that augments an existing calendar-driven or alarm-driven service model.

9. The system of claim 1 , further comprising:

a database configured to store a plurality of measurements, coefficients, and control data associated with a plurality of UPS units,

wherein the central control circuit is further configured to:

execute a machine learning algorithm stored in a memory;

cause the machine learning algorithm to be executed with the plurality of measurements, coefficients, and control data as input; and

cause the machine learning algorithm to output the trained machine learning model trained to automatically detect the degradation of the plurality of capacitor banks.

10. A method for operating uninterruptible power supplies (UPSs), the method comprising:

causing, by a UPS control circuit coupled to a plurality of capacitor banks, an output of a clean electrical power, wherein the clean electrical power corresponds to an electrical power having voltage fluctuations smoothed out;

sampling, by the UPS control circuit, the clean electrical power at a predetermined sample rate;

calculating, by the UPS control circuit, control algorithm outputs based on the sampled clean electrical power;

outputting, by the UPS control circuit, control algorithm results corresponding to the calculated control algorithm outputs to a central control circuit, wherein the central control circuit is remote and distinct from the UPS control circuit;

detecting, by the central control circuit executing a trained machine learning model, a degradation of the plurality of capacitor banks beyond a health threshold based on the control algorithm results such that the trained machine learning model detects the degradation as a function of at least a number of events when a reverse voltage is applied on the plurality of capacitor banks and a number of events when an excessive voltage is applied on the plurality of capacitor banks; and

transmitting, by the central control circuit, a service alert in response to the detection of the degradation of the plurality of capacitor banks.

11. The method of claim 10 , wherein the detection of the degradation is further the function of charge and discharge cycles of the plurality of capacitor banks, charge and discharge durations of the plurality of capacitor banks, capacitor operating voltage, a ripple current value, an internal control coefficient, capacitance value, ambient temperature, capacitor temperature, and an internal coefficient.

12. The method of claim 10 , wherein the clean electrical power comprises a three-phase power.

13. The method of claim 10 , wherein the predetermined sample rate comprises a switching frequency greater than 5 kilohertz (kHz).

14. The method of claim 10 , further comprising:

receiving, by the central control circuit, each of a plurality of measurements, coefficients, and control data from each UPS unit of a plurality UPS units at a particular time; and

causing, by the central control circuit, a database to store the plurality of measurements, coefficients, and control data,

wherein each UPS unit comprises the plurality of capacitor banks and the UPS control circuit, and wherein the central control circuit is distinct and remote from the plurality of UPS units.

15. The method of claim 10 , wherein the control algorithm results comprise capacitor operating voltages associated with each UPS unit of a plurality of UPS units at half load.

16. The method of claim 10 , wherein the control algorithm results comprise ripple current values associated with each UPS unit of a plurality of UPS units measured at an output filter of the UPS unit prior to a receipt of an alternating current (AC) output at a load.

17. The method of claim 10 , further comprising displaying, by a display unit, the service alert prompting a user to provide a maintenance service to the plurality of capacitor banks augmenting an existing calendar-driven or alarm-driven service model.

18. The method of claim 10 , further comprising:

storing, by a database, a plurality of measurements, coefficients, and control data associated with a plurality of UPS units;

executing, by the central control circuit, a machine learning algorithm stored in a memory;

causing, by the central control circuit, the machine learning algorithm to be executed with the plurality of measurements, coefficients, and control data as input; and

causing, by the central control circuit, the machine learning algorithm to output the trained machine learning model trained to automatically detect the degradation of the plurality of capacitor banks.