Real-time estimation of contributions from classes of energy generators in residual load signals

1. A method for managing energy generation in a power generation and distribution system, the method comprising:

obtaining a set of power flow signals measured at a plurality of components within the power generation and distribution system, the set of power flow signals measuring a power consumption and a power generation of the plurality of components in the power generation and distribution system;

calculating residual load data for a given period of time, the residual load data being based on the set of power flow signals to determine a total of the power consumption and a total power generation of the plurality of components within the power generation and distribution system over the given period of time;

optimizing a physical model and a generalized additive model according to an objective function utilizing the residual load data, the physical model defining a relationship between at least one energy generator class and an amount of power generated by the at least one energy generator class, the generalized additive model defining a relationship between a subset of energy generator classes and an amount of power generated by the subset of energy generator classes, the optimizing comprises:

constructing a joint model based on the physical model and the general additive model, and a set of joining parameters based on the physical model; and

performing until convergence:

optimizing the physical model according to the objective function with a set of parameters in the generalized additive model and the set of joining parameters being fixed; and

optimizing the generalized additive model and the set of joining parameters according to the objection function with a set of parameters in the physical model being fixed;

analyzing the residual load data using the physical model after optimization to determine an amount of power of the set of power flow signals attributed to one or more energy generators classified by the at least one energy generator class;

analyzing the residual load data using the generalized additive model after optimization to determine an amount of power of the set of power flow signals attributed to one or more energy generators classified by the subset of energy generator classes, the subset of energy generator classes excluding the at least one energy generator class;

identifying one or more power generators within the power generation and distribution system based on the at least one energy generator class;

detecting an amount of operational power generation from the identified one or more power generators;

calculating one of a power surplus and a power deficiency of the power generation and distribution system based on at least the amount of power of the set of power flow signals attributed to one or more energy generators classified by the at least one energy generator class; and

configuring, based on the calculation of the one of the power surplus and the power deficiency, the one or more energy generators within the power generation and distribution system to increase or decrease energy production.

2. The method of claim 1 , further comprising:

determining, based on the analysis, a total amount of power consumed for the given period of time.

3. The method of claim 1 , wherein the at least one energy generator class includes the one or more power generators within the power generation distribution system.

4. The method of claim 1 , wherein the set of joining parameters in the physical model after optimization is outputted as a physical property of the at least one energy generator class.

5. The method of claim 1 , wherein the at least one energy generator class does not include the one or more power generators within the power generation distribution system.

6. An information processing system for managing energy generation in a power generation and distribution system, the information processing system comprising:

a memory;

at least one processor in communication with the memory, the memory including instructions to configure the at least one processor to:

obtain a set of power flow signals measured at a plurality of components within the power generation and distribution system, the set of power flow signals measures a power consumption and a power generation of the plurality of components within the power generation and distribution system;

calculate residual load data for a given period of time, the residual load data being based on the set of power flow signals to determine a total effect of the power consumption and the power generation over the given period of time;

optimizing a physical model and a generalized additive model according to an objective function utilizing the residual load data, the physical model defining a relationship between at least one energy generator class and an amount of power generated by the at least one energy generator class, the generalized additive model defining a relationship between a subset of energy generator classes and an amount of power generated by the subset of energy generator subclasses, the optimizing comprises:

constructing a joint model based on the physical model and the general additive model, and a set of joining parameters based on the physical model; and

performing until convergence:

optimizing the physical model according to the objective function with a set of parameters in the generalized additive model and the set of joining parameters being fixed; and

optimizing the generalized additive model and the set of joining parameters according to the objection function with a set of parameters in the physical model being fixed;

analyze the residual load data using a physical model after optimization, to determine an amount of power of the set of power flow signals attributed to one or more generators classified by the at least one energy generator class;

analyze the residual load data using the generalized additive model after optimization, to determine an amount of power of the set of power flow signals attributed to one or more energy generators classified by the subset of energy generator classes, the subset of energy generator subclasses excluding the at least one energy generator class;

identify one or more power generators within the power generation and distribution system based on the at least one energy generator class;

detect an amount of operational power generation from the identified one or more power generators;

calculate one of a power surplus and a power deficiency of the power generation and distribution system based on at least the amount of power of the set of power flow signals attributed to one or more energy generators classified by the at least one energy generator class; and

configure, based on the calculation of the one of the power surplus and the power deficiency, the one or more energy generators within the power generation and distribution system to increase or decrease energy production.

7. The method of claim 6 , wherein the set of joining parameters in the physical model after optimization is outputted as a physical property of the at least one energy generator class.

8. A computer program product for managing energy generation in a power generation and distribution system, the computer program product comprising:

a storage medium readable by a processing circuit and storing instructions for execution by the processing circuit for performing a method comprising:

obtaining a set of power flow signals measured at a plurality components within the power generation and distribution system, the set of power flow signals measuring the power consumption and power generation of the plurality of components in the power generation and distribution system;

calculating the residual load data for a given period of time, the residual load data being based on the set of power flow signals to determine a total effect of the power consumption and power generation of the plurality of components within the power generation and distribution system over the given period of time;

optimizing a physical model and a generalized additive model according to an objective function utilizing the residual load data, the physical model defining a relationship between at least one energy generator class and an amount of power generated by the at least one energy generator class, the generalized additive model defining a relationship between a subset of energy generator classes and an amount of power generated by the subset of energy generator classes the optimizing comprises;

constructing a joint model based on the physical model and the generalized additive model, and a set of joining parameters based on the physical model; and

performing until convergence:

optimizing the physical model according to the objective function with a set of parameters in the generalized additive model and the set of joining parameters being fixed; and

optimizing the generalized additive and the set of joining parameters according to the objection function with a set of parameters in the physical model being fixed;

analyzing the residual load data using a physical model after optimization to determine an amount of power of the set of power flow signals attributed to one or more energy generators classified by the at least one energy generator class;

analyzing the residual load data using the generalized additive model after optimization to determine an amount of power of the set of power flow signals attributed to one or more energy generators classified by the subset of energy generator classes, the subset of energy generator classes excluding the at least one energy generator class;

identifying one or more power generators within the power generation and distribution system based on the at least one energy generator class;

detecting an amount of operational power generation from the identified one or more power generators;

calculating one of a power surplus and a power deficiency of the power generation and distribution system based on at least the amount of power of the set of power flow signals attributed to one or more energy generators classified by the at least one energy generator class; and

configuring, based on the calculation of the one of the power surplus and the power deficiency, the one or more energy generators within the power generation and distribution system to increase to decrease energy production.

9. The computer program product of claim 8 , wherein the set of joining parameters in the physical model after optimization is outputted as a physical property of the at least one energy generator class.