Automatic system for monitoring electric power consumption

Patent metadata

Field	Value
Publication number	US-10333304-B2
Application number	US-201415525621-A
Country	US
Kind code	B2
Filing date	Nov 18, 2014
Priority date	Nov 18, 2014
Publication date	Jun 25, 2019
Grant date	Jun 25, 2019

Abstract

A system for managing devices supplied through a power grid is proposed. Each device is configured to consume over time electric power according to at least one respective power profile when operating. The system comprises at least one unit interfaced with the devices for exchanging data. The at least one unit collects power profile data indicative of the power profiles of the devices, and generates a time schedule of the device operations by distributing over time start times of the power profiles in such a way that at any time the total power consumption of the devices is kept under a maximum power threshold of the power grid.

First claim

The invention claimed is:

1. A system for managing devices supplied through a power grid, each device being configured to consume over time electric power according to at least one respective power profile when operating, the system comprising at least one unit interfaced with the devices for exchanging data, the at least one unit being configured to:
collect power profile data indicative of the power profiles of said devices;
generate a time schedule of the device operations by distributing over time start times of the power profiles in such a way that at any time the total power consumption of the devices is kept under a maximum power threshold of the power grid, wherein:
the at least one unit is configured to generate said time schedule through a particle swarm optimization approach by:
a) randomly generating a population of candidate time schedules within a search space comprising a set of the possible candidate time schedules;
b) subdividing the population of candidate time schedules in a first group of candidate time schedules and in a second group of candidate time schedules;
c) performing phases d) and e) at least once;
d) moving the candidate time schedules of the first group in the search space following a random walk based on a stable probability distribution having an infinite variance;
e) moving the candidate time schedules of the second group in the search space following a random walk based on a probability distribution having a finite variance;
f) selecting a candidate time schedule based on a cost of said candidate time schedule corresponding to its position in the search space calculated by means of a cost function.

2. The system of claim 1 , wherein:
said random walk based on a stable probability distribution having an infinite variance is a random walk based on a Lévy probability distribution, and
said random walk based on a probability distribution having a finite variance is a uniform random walk.

3. The system of claim 1 , wherein the at least one unit is further configured to:
for each candidate time schedule of the first and the second groups, moving said candidate time schedule by:
a) calculating a corresponding new position in the search space for said candidate time schedule,
b) moving said candidate time schedule to the corresponding new position conditioned to a comparison between a cost of said candidate time schedule corresponding to the current position in the search space calculated by means of the cost function and a cost of said candidate time schedule corresponding to the new position in the search space calculated by means of the cost function.

4. The system of claim 3 , wherein the at least one unit is further configured to:
select a candidate time schedule best position in the search space among positions previously taken by all the candidate time schedules in the search space based on the cost thereof, and
move each candidate time schedule by:
if the candidate time schedule belongs to the first group, calculating said corresponding new position based on a first difference between the best position and the current position of the candidate time schedule;
if the candidate time schedule belongs to the second group, calculating said corresponding new position based on a second difference between the current positions of two random candidate time schedules of the population.

5. The system of claim 4 , wherein the at least one unit is further configured to move each candidate time schedule by:
if the candidate time schedule belongs to the first group, calculating said corresponding new position by varying the current position thereof by a first term based on a product between said first difference and a random number with Lévy distribution;
if the candidate time schedule belongs to the second group, calculating said corresponding new position by varying the current position thereof by a second term based on a product between said second difference and a random number with a uniform distribution.

6. The system of claim 4 , wherein the at least one unit is further configured to move said candidate time schedule to the corresponding new position if the cost of the candidate time schedule corresponding to the current position is higher than the cost of the candidate time schedule corresponding to the new position.

7. The system of claim 6 , wherein the at least one unit is further configured to move said candidate time schedule to the corresponding new position when the cost of the candidate time schedule corresponding to the current position is lower than the cost of the candidate time schedule corresponding to the new position conditioned to a parameter corresponding to a ratio of a third term and a fourth term, the third term corresponding to the difference between the cost of the candidate time schedule corresponding to the new position and the cost of the candidate time schedule in the candidate time schedule best position, the fourth term corresponding to the distance between the new position and the candidate time schedule best position.

8. The system of claim 1 , wherein the cost function of a candidate time schedule depends on the difference between the maximum power threshold of the power grid and a power consumption of the devices that would occur if the devices operated following the candidate time schedule.

9. The system of claim 1 , wherein the at least one unit comprises a control unit configured to automatically control the operation of the devices based on the time schedule.

10. The system of claim 1 , wherein the at least one unit comprises a user equipment configured to allow a user to view details of the time schedule.

11. A method for managing devices supplied through a power grid, each device being configured to consume over time electric power according to at least one respective power profile when operating, the method comprising:
collecting power profile data indicative of the power profiles of said devices;
generating a time schedule of the device operations by distributing over time start times of the power profiles in such a way that at any time the total power consumption of the devices is kept under a maximum power threshold of the power grid, wherein said generating a time schedule comprises generating said time schedule through a swarm optimization approach by performing the following phases:
a) randomly generating a population of candidate time schedules within a search space comprising a set of the possible candidate time schedules;
b) subdividing the population of candidate time schedules in a first group of candidate time schedules and in a second group of candidate time schedules;
c) performing phases d) and e) at least once;
d) moving the candidate time schedules of the first group in the search space following a random walk based on a stable probability distribution having an infinite variance;
e) moving the candidate time schedules of the second group in the search space following a random walk based on a probability distribution having a finite variance;
f) selecting a candidate time schedule based on a cost of said candidate time schedule corresponding to its position in the search space calculated by means of a cost function.

12. The method of claim 11 , wherein:
said random walk based on a stable probability distribution having an infinite variance is a random walk based on a Lévy probability distribution, and
said random walk based on a probability distribution having a finite variance is a uniform random walk.

13. The method of claim 11 , wherein the method further comprises:
for each candidate time schedule of the first and the second groups, mo

Assignees

• Telecom Italia Spa

Inventors

• Borean Claudio
• Grasso Ennio

Classifications

• H02J2105/42

  Home appliances · CPC title

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• Y02B70/3225

  Demand response systems, e.g. load shedding, peak shaving · CPC title

• Y04S20/222

  Demand response systems, e.g. load shedding, peak shaving · CPC title

• G01R22/063

  related to remote communication · CPC title