Energy management system with machine learning
An energy management system including a universal energy flow manager that has a housing, an energy storage device disposed in the housing, a power electronics module disposed in the housing and adapted to convert and manage power and a connections interface. It also includes distribution and communications module with a HVDC Bus (High-Voltage Direct Current Bus) having variable power limits that powers an entire load requirement of one or more coupled electrical loads up to a defined power limit determined by an aggregation of power from one or more coupled energy sources. The one or more coupled electrical loads and the one or more coupled energy sources are external to the universal energy flow manager and connect to the HVDC Bus through the connections interface.
1. A computer-implemented method comprising:
receiving an energy demand state of one or more coupled electrical loads coupled to a universal energy flow manager via a load application specific hardware (load ASH), the energy demand state being indicative of a desired amount of energy needed by the one or more coupled electrical loads in an energy management environment;
receiving an available energy state of one or more coupled energy sources coupled to the universal energy flow manager via a source application specific hardware (source ASH), the available energy state being indicative of an available amount of energy from the one or more coupled energy sources in the energy management environment;
generating input data using at least the energy demand and available energy states;
extracting one or more features from the input data, the one or more features representative of a characteristic of a request for a power delivery proposal operation,
proposing, using a power delivery module, at least one power delivery proposal for the one or more coupled electrical loads;
and performing a power delivery operation based on the power delivery proposal.
2. The computer-implemented method of claim 1 , wherein the power delivery module is a machine learning engine.
3. The computer-implemented method of claim 1 , further comprising:
performing the power delivery operation through a load ASH of the one or more coupled electrical loads.
4. The computer-implemented method of claim 3 , wherein the power delivery operation is performed automatically.
5. The computer-implemented method of claim 1 , further comprising:
generating, by attributes prioritization, a set of attributes of the energy management environment to enforce; and
proposing the at least one power delivery proposal based on one or more attributes of the set of attributes.
6. The computer-implemented method of claim 5 , wherein the attributes include an attribute selected from the list consisting of minimizing utility cost, minimizing charging time, maximizing availability of backup energy, maximize excess energy sales, and maximizing battery life.
7. The computer-implemented method of claim 1 , wherein the power delivery proposal comprises instructions to perform a DC fast charge or one or more electric vehicles.
8. The computer-implemented method of claim 1 , wherein the power delivery proposal comprises instructions to maximize use of renewable energy sources during a defined time interval.
9. The computer-implemented method of claim 1 , wherein the power delivery proposal is provided in real time.
10. The computer-implemented method of claim 1 , wherein the power delivery proposal comprises load shedding instructions.
11. The computer-implemented method of claim 1 , wherein the power delivery proposal comprises powering a plurality of loads by time division multiplexing of load source pairs.
12. The computer-implemented method of claim 1 , wherein the input data further comprise data selected from the list consisting of historical device use data, a fast charging requirement, a weather forecast, a calendar data, a current electricity consumption demand, a vehicle energy demand profile, a home energy demand profile, and a battery lifetime.
13. The computer-implemented method of claim 1 , further comprising:
providing feedback for the power delivery module indicative of an accuracy of proposals to reinforce the power delivery module.
14. The computer-implemented method of claim 1 , further comprising:
providing the power delivery proposal in real time.
15. A computer system comprising a processor configured to:
receive an energy demand state of one or more coupled electrical loads coupled to a universal energy flow manager via a load application specific hardware (load ASH), the energy demand state being indicative of a desired amount of energy needed by the one or more coupled electrical loads in an energy management environment;
receive an available energy state of one or more coupled energy sources coupled to the universal energy flow manager via a source application specific hardware (source ASH), the available energy state being indicative of an available amount of energy from the one or more coupled energy sources in the energy management environment;
generate input data using at least the energy demand and available energy states;
extract one or more features from the input data, the one or more features representative of a characteristic of a request for a power delivery proposal operation,
propose, using a power delivery module, at least one power delivery proposal for the one or more coupled electrical loads;
and perform a power delivery operation based on the power delivery proposal.
16. A non-transitory computer-readable storage medium storing a program which, when executed by a computer system, causes the computer system to:
receive an energy demand state of one or more coupled electrical loads coupled to a universal energy flow manager via a load application specific hardware (load ASH), the energy demand state being indicative of a desired amount of energy needed by the one or more coupled electrical loads in an energy management environment;
receive an available energy state of one or more coupled energy sources coupled to the universal energy flow manager via a source application specific hardware (source ASH), the available energy state being indicative of an available amount of energy from the one ore one or more coupled energy sources in the energy management environment;
generate input data using at least the energy demand and available energy states;
extract one or more features from the input data, the one or more features representative of a characteristic of a request for a power delivery proposal operation,
propose, using a power delivery module, at least one power delivery proposal for the one or more coupled electrical loads;
and perform a power delivery operation based on the power delivery proposal.