EP3832619B1 - Methods and devices for monitoring lifestyle changes - Google Patents

Description

Technical area

The invention relates to the field of data analysis, in particular the extraction of intelligent information from rich and noisy information media.

More particularly, the invention relates to methods for monitoring an evolution over time of at least one activity of a person, computer programs for implementing such methods, data storage devices storing such programs and processing circuits including such data storage devices.

Prior art

We know of US 4,990,893 a method for an alarm system based on a recording of energy consumption. We know of US 2016/0371593 a device and a daily activity inference program corresponding to the preamble of claim 1. Currently, in the field of assistance to the elderly, numerous solutions are developed to offer services detecting sudden events, such as a fall or discomfort, and which require an immediate reaction.

These solutions rely on physical sensors configured to detect these sudden events.

According to some solutions, physical sensors are installed in the home. For example, these physical sensors can be integrated into connected objects, installed in the home by defining a precise communication protocol. Over time, it is possible that hardware and/or software modifications to connected objects or network equipment may affect this communication.

According to other solutions, the physical sensors are worn directly by the user. Wearable sensors are invasive and are not always accepted because they are visible and restrictive.

Furthermore, whatever the solution chosen, the proper functioning of the sensors as well as their power supply must be managed, which requires not only annoying maintenance for the user but also a risk of breakdown.

Monitoring slow developments over a long period, for example of a few weeks, months or years, is therefore difficult.

Summary

This disclosure improves the situation.

1. a) obtenir, pour une période temporelle subdivisée en une pluralité d'intervalles de temps consécutifs, une série temporelle comprenant, pour chaque intervalle de temps, une valeur, associée audit intervalle de temps, indiquant un niveau de consommation d'au moins un équipement électrique de l'installation électrique au cours dudit intervalle de temps,
2. b) pour chaque intervalle de temps, comparer la valeur associée audit intervalle de temps avec au moins une valeur de référence,

le procédé comprenant en outre, pour au moins une activité de la personne:

• c) sur la base des comparaisons, déterminer une information ciblée indicative de ladite activité de la personne au cours de la période temporelle, et
• d) surveiller une évolution au cours du temps de ladite activité de la personne sur la base de ladite information ciblée et d'au moins une information ciblée de référence indicative d'une activité habituelle de la personne au cours d'une pluralité de périodes temporelles de référence précédant ladite période temporelle.

A method is proposed for monitoring an evolution over time of at least one activity of a person in a home equipped with an electrical installation comprising at least one piece of electrical equipment, the method comprising:

1. a) obtain, for a time period subdivided into a plurality of consecutive time intervals, a time series comprising, for each time interval, a value, associated with said time interval, indicating a level of consumption of at least one piece of equipment electrical of the electrical installation during said time interval,
2. b) for each time interval, compare the value associated with said time interval with at least one reference value,

the method further comprising, for at least one activity of the person:

• c) based on the comparisons, determine targeted information indicative of said activity of the person during the time period, and
• d) monitoring an evolution over time of said activity of the person on the basis of said targeted information and at least one targeted reference information indicative of a usual activity of the person during a plurality of temporal periods reference preceding said time period.

By “activity” is meant an activity of finite duration and of a habitual nature, such as a period of sleep, a period of absence, preparation of meals, use of domestic hot water, etc.

By “targeted information indicative of said activity” is meant information indicative of such an activity, and therefore generally of the person's pace of life.

The electrical installation is for example domestic and may include a plurality of electrical equipment such as lamps, heating equipment, household appliances, IT, multimedia, etc.

The electrical installation is capable in particular of drawing electrical energy from an electrical network to ensure the operation of the electrical equipment(s).

• maximiser la richesse des informations formant la série temporelle en maximisant la fréquence de mesure, ce qui équivaut à minimiser la durée des intervalles de temps consécutifs,
• minimiser les ressources matérielles nécessaires pour stocker et manipuler ces informations, ce qui équivaut à minimiser le nombre de points de mesure par série temporelle, en minimisant la fréquence de mesure et la durée de la période temporelle,
• minimiser l'impact de la transmission de ces informations à travers un réseau de communication, ce qui implique de maximiser le nombre de points de mesure par série temporelle en maximisant la durée de la période temporelle, et
• maximiser la réactivité de la surveillance, ce qui implique de fixer la durée de la période temporelle en deçà d'un seuil maximal.

The duration of the “time period” and “consecutive time intervals” may be guided by different objectives, and may in particular represent a compromise between the following objectives:

• maximize the wealth of information forming the time series by maximizing the measurement frequency, which is equivalent to minimizing the duration of consecutive time intervals,
• minimize the hardware resources required to store and manipulate this information, which is equivalent to minimizing the number of measurement points per time series, by minimizing the measurement frequency and the duration of the time period,
• minimize the impact of transmitting this information through a communications network, which involves maximizing the number of measurement points per time series by maximizing the duration of the time period, and
• maximize the responsiveness of monitoring, which involves setting the duration of the time period below a maximum threshold.

As an indication, the “time period” can for example be of the order of a few hours or days and be subdivided into “consecutive time intervals” of the order of ten seconds, a few minutes, tens of minutes or hours.

The “time series” defines the evolution of electrical energy consumption by at least one piece of electrical equipment in the electrical installation during the time period.

Here, “at least one piece of equipment” can designate all of the equipment in the electrical installation during the time period.

For example, the power withdrawn by the electrical installation during a time interval can be measured using an electric meter which records, for example, a time-stamped average value of the electrical power withdrawn over an equal integration period. at the time interval. Thus, according to this example, the meter readings can form a load curve of the electrical installation over the time period.

This load curve is a set of time-stamped average values of electrical power drawn by the electrical installation, or “consumption levels”, over “consecutive time intervals” each corresponding to an integration period.

Alternatively, the electrical installation may include a plurality of electrical equipment, and the “time series” may define the evolution of electrical energy consumption by only part of the equipment in the electrical installation during the time period.

For example, this part can designate a single piece of equipment, or a group of pieces of equipment of the same type, or a group of pieces of equipment located in the same room of the home.

In this example, the “time series” can be obtained in different ways, and can in particular be extracted from a load curve of the electrical installation during the time period as defined above.

Thanks to the method according to the invention, it is possible to monitor an evolution over time of a person's activity over a long period, for example of the order of a few weeks, months or years.

In addition, this monitoring is carried out by overcoming the disadvantages mentioned above associated with the use of physical sensors as well as their wearing by the person or their installation in the person's home.

According to another aspect, a computer program is proposed comprising instructions for implementing a method as described above when this program is executed by a processor.

According to another aspect, a data storage device is proposed, or a non-transitory recording medium readable by a computer, on which a program is recorded for the implementation of a method as described above when this program is executed by a processor.

According to another aspect, a processing circuit is proposed comprising a processor connected to a non-transitory recording medium as described above.

The process as described above also provides the following characteristics:

• l'information ciblée indicative de ladite présence ou absence de la personne est une information ciblée du type d'un indicateur de présence,
• l'au moins une valeur de référence comprend une valeur de consommation électrique passive,
  le procédé comprenant :
  • b1) pour chaque intervalle de temps, comparer la valeur associée audit intervalle de temps avec la valeur de consommation électrique passive, et
  • c1) pour ladite présence ou absence de la personne, sur la base des comparaisons, déterminer l'information ciblée du type d'un indicateur de présence de manière à indiquer la période temporelle comme étant:
• une période d'absence de la personne si chaque valeur associée à un intervalle de temps est inférieure ou égale à la valeur de consommation électrique passive, ou
• une période de présence de la personne au cours de la période temporelle si au moins une valeur associée à un intervalle de temps est supérieure à la valeur de consommation électrique passive.

- at least one activity of the person is a presence or absence of the person during at least part of the time period,

• the targeted information indicative of said presence or absence of the person is targeted information of the type of a presence indicator,
• the at least one reference value comprises a passive electrical consumption value,
  the process comprising:
  • b1) for each time interval, compare the value associated with said time interval with the passive electricity consumption value, and
  • c1) for said presence or absence of the person, on the basis of the comparisons, determine the targeted information of the type of a presence indicator so as to indicate the temporal period as being:
• a period of absence of the person if each value associated with a time interval is less than or equal to the passive electricity consumption value, or
• a period of presence of the person during the temporal period if at least a value associated with a time interval is greater than the passive power consumption value.

Thanks to the characteristics described above, it is possible to detect the presence or absence of a person in a home during a time period such as a day.

It is also possible to monitor the alternation of periods of presence and absence in order to identify, for example, if a period of presence or absence is unusually long.

The passive electricity consumption of electrical equipment corresponds to the electricity consumption of this equipment in the absence of any human activity causing electricity consumption. For example, the electrical equipment may be a television whose passive electrical consumption corresponds to its consumption in the standby state. Certain electrical equipment, for example a refrigerator, has a practically constant electricity consumption independent of any human activity over a long period of time. This electricity consumption corresponds to their passive electricity consumption.

The passive electricity consumption of an electrical installation corresponds to the sum of the passive electricity consumption of each electrical equipment in the installation. The passive electrical consumption of an electrical installation can be obtained or determined in several ways, for example by accessing a database of passive electrical consumption.

• a1) obtenir, pour une période temporelle historique subdivisée en une pluralité d'intervalles de temps historiques consécutifs, une série temporelle historique comprenant, pour chaque intervalle de temps historique, une valeur historique, associée audit intervalle de temps historique, indiquant un niveau de consommation d'au moins un équipement électrique de l'installation électrique au cours dudit intervalle de temps historique, et
• a2) estimer, sur la base de la série temporelle historique, la valeur de consommation électrique passive dudit au moins un équipement électrique de l'installation électrique.

The characteristics set out in this paragraph and the following can, optionally, be implemented. They can be implemented independently of each other or in combination with each other. In one embodiment, the method further comprises:

• a1) obtain, for a historical time period subdivided into a plurality of consecutive historical time intervals, a historical time series comprising, for each historical time interval, a historical value, associated with said historical time interval, indicating a level of consumption at least one piece of electrical equipment of the electrical installation during said historical time interval, and
• a2) estimate, based on the historical time series, the value of passive electrical consumption of said at least one piece of electrical equipment of the electrical installation.

For a given piece of equipment or a given installation, the historical electrical consumption values of said equipment or said installation are a reliable source for estimating the passive electrical consumption used, for example, to determine the presence indicator of the person as defined above.

The estimation of passive electricity consumption based on the historical time series can be implemented using different statistical methods. For example, passive power consumption can be defined as the first quartile of historical values, or the first decile, or calculated from the mean and standard deviation of historical values, etc.

In one embodiment, estimating, on the basis of the historical time series, the passive electrical consumption value of said at least one piece of electrical equipment of the electrical installation comprises:
a3) implement a one-factor statistical test between a first group formed of historical values less than or equal to a threshold and a second group formed of historical values greater than said threshold, the result of the statistical test being the estimated value of passive electrical consumption said at least one piece of electrical equipment of the electrical installation.

In one embodiment, the statistical test is an ANOVA test.

• les valeurs historiques inférieures ou égales au seuil de rupture correspondent à des fluctuations de consommation électrique au cours du temps, non imputables à une action délibérée de la personne, et
• les valeurs historiques supérieures au seuil de rupture correspondent à une consommation électrique nécessairement associée à une activité de la personne.

Thus, it is possible to characterize the passive electrical consumption value as representing a breaking threshold among the historical values, so that:

• historical values less than or equal to the breakdown threshold correspond to fluctuations in electricity consumption over time, not attributable to a deliberate action of the person, and
• historical values above the breakdown threshold correspond to electrical consumption necessarily associated with a person's activity.

In this way, it is possible to determine, for each time interval, whether the electricity consumption during said time interval is associated or not with an activity of the person and thus to monitor the evolution of the activity of the person during each time interval.

• la période temporelle est indiquée comme étant une période de présence de la personne,
• au moins une activité de la personne est une heure de coucher de la personne,
• l'information ciblée indicative de ladite heure de coucher de la personne est une information temporelle du type d'une heure de coucher, et

le procédé comprend :
c2) sur la base des comparaisons, déterminer l'information temporelle du type d'une heure de coucher, ladite information temporelle étant associée au dernier intervalle de temps de la période temporelle à compter d'une heure prédéterminée tel que la valeur associée audit intervalle de temps est supérieure à la valeur de consommation électrique passive.In one embodiment,

• the time period is indicated as being a period of presence of the person,
• at least one activity of the person is a bedtime of the person,
• the targeted information indicative of said bedtime of the person is temporal information of the type of a bedtime, and

the process includes:
c2) on the basis of the comparisons, determine the temporal information of the type of a bedtime, said temporal information being associated with the last time interval of the temporal period from a predetermined time such as the value associated with said interval of time is greater than the passive electricity consumption value.

Thus, it is possible to monitor, on the basis of monitoring the person's electricity consumption over time excluding periods of absence of the person, the evolution over time of the bedtime of the person. person.

• la période temporelle est indiquée comme étant une période de présence de la personne,
• au moins une activité de la personne est une heure de lever de la personne,
• l'information ciblée indicative de ladite heure de lever de la personne est une information temporelle du type d'une heure de lever, et

le procédé comprend :
c3) sur la base des comparaisons, déterminer l'information temporelle du type d'une heure de lever, ladite information temporelle étant associée au premier intervalle de temps de la période temporelle à compter d'une heure prédéterminée tel que la valeur associée audit intervalle de temps est supérieure à la valeur de consommation électrique passive.In one embodiment,

• the time period is indicated as being a period of presence of the person,
• at least one activity of the person is an hour of waking up of the person,
• the targeted information indicative of said waking time of the person is temporal information of the type of a waking time, and

the process includes:
c3) on the basis of the comparisons, determine the temporal information of the type of a sunrise time, said temporal information being associated with the first time interval of the temporal period from a predetermined time such as the value associated with said interval of time is greater than the passive electricity consumption value.

Thus, it is possible to monitor, on the basis of monitoring the person's electricity consumption over time excluding periods of absence of the person, the evolution over time of the time of getting up from the bed. person.

Time indications such as a person's waking time and bedtime allow the person's circadian rhythm to be monitored.

• l'installation électrique comprend au moins un équipement électrique de cuisson,
• l'activité de la personne est une occurrence de prise de repas par la personne,
• pour chacun des intervalles de temps consécutifs, le niveau de consommation électrique au cours dudit intervalle de temps est un niveau de consommation électrique de l'installation électrique au cours dudit intervalle de temps,
• l'au moins un niveau de consommation électrique de référence comprend un seuil de consommation électrique de l'installation électrique, indicatif d'un fonctionnement d'au moins un équipement électrique de cuisson, et

le procédé comprend :

• b2) pour chaque intervalle de temps, comparer le niveau de consommation électrique de l'installation électrique au cours dudit intervalle de temps avec ledit seuil de consommation, et
• c4) sur la base des comparaisons des niveaux de consommation électrique de l'installation électrique avec ledit seuil de consommation, déterminer une information ciblée du type d'un indicateur de prise de repas d'une personne au cours d'au moins une partie de la période temporelle.

In one embodiment,

• the electrical installation includes at least one piece of electrical cooking equipment,
• the person's activity is an occurrence of the person taking a meal,
• for each of the consecutive time intervals, the level of electricity consumption during said time interval is a level of electricity consumption of the electrical installation during said time interval,
• the at least one reference electrical consumption level comprises an electrical consumption threshold of the electrical installation, indicative of operation of at least one piece of electrical cooking equipment, and

the process includes:

• b2) for each time interval, compare the level of electrical consumption of the electrical installation during said time interval with said consumption threshold, and
• c4) on the basis of comparisons of the electrical consumption levels of the electrical installation with said consumption threshold, determine targeted information of the type of an indicator of meal intake of a person during at least part of the time period.

Thus, it is possible to detect occurrences and meal time slots in order to monitor the person's eating frequency over time.

• l'installation électrique comprend un chauffe-eau électrique,
• l'activité de la personne est une consommation d'eau chaude sanitaire issue du chauffe-eau électrique par la personne,
• pour chacun des intervalles de temps consécutifs, le niveau de consommation électrique au cours dudit intervalle de temps est un niveau de consommation électrique du chauffe-eau électrique au cours dudit intervalle de temps,
• l'au moins un niveau de consommation électrique de référence comprend un seuil de fonctionnement du chauffe-eau électrique, et

le procédé comprend :

• b3) pour chaque intervalle de temps, comparer le niveau de consommation électrique dudit chauffe-eau électrique au cours dudit intervalle de temps avec le seuil de fonctionnement dudit chauffe-eau électrique, et
• c5) sur la base des comparaisons, déterminer une information ciblée du type d'un indicateur de consommation d'eau chaude sanitaire issue du chauffe-eau électrique au cours de la période temporelle.

In one embodiment,

• the electrical installation includes an electric water heater,
• the person's activity is consumption of domestic hot water from the electric water heater by the person,
• for each of the consecutive time intervals, the level of electricity consumption during said time interval is a level of electricity consumption of the electric water heater during said time interval,
• the at least one reference electricity consumption level comprises an operating threshold of the electric water heater, and

the process includes:

• b3) for each time interval, compare the level of electricity consumption of said electric water heater during said time interval with the operating threshold of said electric water heater, and
• c5) on the basis of the comparisons, determine targeted information of the type of an indicator of domestic hot water consumption from the electric water heater during the time period.

Thus, it is possible to detect occurrences and time slots of domestic hot water consumption in order to monitor the person's level of hygiene over time.

• a4) obtenir un groupe d'informations ciblées de référence comprenant une pluralité d'informations ciblées de référence du même type que l'information ciblée déterminée, chaque information ciblée de référence dudit type étant indicative d'une activité de la personne au cours d'une période temporelle de référence passée, antérieure à ladite période temporelle,
• a5) obtenir un groupe d'informations ciblées comprenant, outre l'information ciblée déterminée, une pluralité d'informations ciblées précédentes du même type que l'information ciblée déterminée, chaque information ciblée précédente étant indicative de ladite activité de la personne au cours d'une période temporelle passée, antérieure à ladite période temporelle et ultérieure auxdites périodes temporelles de référence passées, et
• surveiller une évolution au cours du temps d'une activité de la personne comprend, pour au moins une information ciblée déterminée :
  d1) surveiller une évolution au cours du temps de ladite information ciblée sur la base dudit groupe d'informations ciblées et dudit groupe d'informations ciblées de référence.

In one embodiment, the method comprises:

• a4) obtaining a group of targeted reference information comprising a plurality of targeted reference information of the same type as the determined targeted information, each targeted reference information of said type being indicative of an activity of the person during a past reference time period, prior to said time period,
• a5) obtaining a group of targeted information comprising, in addition to the determined targeted information, a plurality of previous targeted information of the same type as the determined targeted information, each previous targeted information being indicative of said activity of the person during 'a past time period, prior to said time period and subsequent to said past reference time periods, and
• monitoring an evolution over time of an activity of the person includes, for at least one specific targeted piece of information:
  d1) monitoring an evolution over time of said targeted information on the basis of said group of targeted information and said group of reference targeted information.

In an exemplary embodiment, the targeted information is associated with a time period of 24 hours. The group of targeted information includes six previous targeted information of the type of a person presence indicator. This previous targeted information is associated with six consecutive 24-hour past time periods. Thus, the targeted information group has targeted information for each day of the past week. For example, if the targeted information is of the type of a presence indicator, then the group of targeted information indicates for each day of the past week whether the person is present or not. In addition, the group of targeted information can also include other targeted information such as a waking up time and thus indicate, for each day of presence of the person during the past week, their waking up time.

The reference targeted information group comprises N reference targeted information associated with N past reference time periods. Thus, the group of targeted reference information can have, for each day during the weeks or months preceding the previous week, targeted reference information.

In this example, the process can be repeated each new week, integrating the group of targeted information from the previous week into the group of reference targeted information and acquiring a new group of targeted information for the new week.

Monitoring the evolution of this targeted information over time can aim to detect different types of events.

In one embodiment, monitoring an evolution over time of said targeted information comprises:
d1) determining a difference between, on the one hand, the targeted information of said group of targeted information and, on the other hand, the targeted reference information of said group of targeted reference information, and, if said difference exceeds a threshold of predetermined difference, detect an event indicating a break in the activity of the person.

The break occurs when the targeted information is stable before a very significant evolution over a short period of time of the order of a week then becomes stable again at a different value.

In one embodiment, monitoring an evolution over time of said targeted information comprises:
d2) determine a derivative of said targeted information with respect to time on the basis of the group of targeted information and the group of reference targeted information, and, if said derivative exceeds a predetermined variation threshold, detect an event indicating a change in the person's activity.

Slow evolution occurs when the targeted information, after having remained stable, begins to drift continuously over a long period of time of the order of several weeks.

By monitoring the targeted information over time according to one of the preceding embodiments, it is possible to detect with each new added week a break or a significant slow evolution.

In one embodiment, the method further comprises, after monitoring an evolution over time of said targeted information and detecting an event,
e) generate an alert in response to detection of the event.

Brief description of the drawings

• Fig. 1
  La [Fig. 1] illustre schématiquement la structure d'un circuit de traitement, dans un exemple de réalisation, pour la mise en oeuvre du procédé proposé.
• Fig. 2 à 5
  Chacune des [Fig. 2], [Fig. 3], [Fig. 4] et [Fig. 5] illustre une différente forme d'ordinogramme d'un algorithme général d'un programme informatique, pour différents exemples particuliers de réalisations, pour la mise en oeuvre du procédé proposé.

Other characteristics, details and advantages will appear on reading the detailed description below, and on analyzing the appended drawings, in which:

• Fig. 1
  There [ Fig. 1 ] schematically illustrates the structure of a processing circuit, in an exemplary embodiment, for implementing the proposed method.
• Fig. 2 to 5
  Each of the [ Fig. 2 ], [ Fig. 3 ], [ Fig. 4 ] And [ Fig. 5 ] illustrates a different form of flowchart of a general algorithm of a computer program, for different particular examples of implementations, for the implementation of the proposed method.

Description of embodiments

With the emergence of so-called intelligent and communicating electricity meters, it is possible to recover the actual electricity consumption of an electrical installation on a frequent basis. For example, it is possible to obtain indicative values of the electricity consumption of the installation electrically periodically, with a period which may be a few seconds, 10 minutes, 30 minutes or 1 hour.

In the examples described below, we consider a domestic electrical installation located in a home where an elderly person lives alone.

The domestic electrical installation taken as an example is equipped with an electricity meter which records and stores locally, in periods of 30 minutes, the electrical consumption of the electrical installation over the last 30 minutes. Thus, every 30 minutes, a value is acquired, corresponding to the average power withdrawn from the electrical network during the last 30-minute time interval.

Every 24 hours, a load curve, formed from the last 48 values acquired, is thus created and transmitted to a remote server including a processing circuit.

• à un support d'enregistrement non transitoire MEM 200 sur lequel est enregistré un programme pour la mise en oeuvre d'un procédé tel que décrit ci-après lorsque ce programme est exécuté par le processeur PROC, et
• à une interface de communication COM 300 permettant au circuit de traitement notamment de recevoir des informations de l'extérieur, par exemple les relevés reçus des compteurs communicants, et de transmettre des informations vers l'extérieur, par exemple des alertes générées à destination de terminaux.

An example of such a processing circuit is shown in [ Fig. 1 ]. The processing circuit shown includes a PROC 100 processor connected:

• to a non-transitory recording medium MEM 200 on which is recorded a program for implementing a method as described below when this program is executed by the processor PROC, and
• to a COM 300 communication interface allowing the processing circuit in particular to receive information from the outside, for example readings received from communicating meters, and to transmit information to the outside, for example alerts generated for terminals .

We now refer to the [ Fig. 2 ] which illustrates an example of carrying out a monitoring method.

The processing circuit obtains OBT HISTO (S11) a consumption history of the electrical installation. This history can be formed from the daily load curves received by the processing circuit via the COM communication interface during previous weeks or previous months.

These load curves are historical data based solely on the habits of the person who lives in their home. These data are collected in a non-invasive manner for the person. Another advantage of collecting these load curves as historical data is that communicating electricity meters are certified equipment, guaranteeing the reliability of the load curves obtained.

The processing circuit estimates EST PASS (S13) a passive electrical consumption of the electrical installation, in other words, when the person has not deliberately activated electrical equipment (for example by lighting hotplates or A television).

Typically, this passive consumption essentially includes the consumption of the refrigerator, the freezer and the standby of the devices (television, box, connected device, etc.) in an example of an electrical installation including the aforementioned equipment.

To estimate this passive consumption, it can be planned to collect all the electricity consumption values over the course of, for example, a week, i.e. 336 values, and to implement TEST STAT (S12) a statistical test on the basis of the values. collected.

• un premier groupe formé des valeurs inférieures ou égales à un seuil de consommation S, et
• un deuxième groupe formé des valeurs supérieures au seuil de consommation S.

The collected values can be divided into:

• a first group formed of values less than or equal to a consumption threshold S, and
• a second group formed of values greater than the consumption threshold S.

This consumption threshold S is considered as a candidate for the solution of an optimization problem, or statistical test, aimed at estimating the level of passive consumption of the installation

To determine this level, it is possible to implement, for example, a one-way ANOVA test between the two aforementioned groups.

Thus, different consumption threshold values S are tested. These different threshold values tested are between the minimum and maximum values collected for electricity consumption of the installation during the week.

For each threshold value tested, the collected consumption values are divided into two groups and the p-value of the ANOVA test is determined. The level of passive consumption is estimated as the threshold value tested which minimizes the p-value of the ANOVA test.

The processing circuit obtains daily OBT SERIE (S10) a load curve as described above.

The processing circuit compares COMP CONSO/PASS (S21) each value of the load curve obtained with the estimated passive consumption level.

From these comparisons, the processing circuit can determine DET INDIC PRES (S31) an indicator of the presence of the person in the home.

Periods of more than 24 hours which do not have half-hourly consumption exceeding the level of passive consumption can in fact be considered as periods of absence. It is also possible to identify in the consumption history the longest periods which meet this condition in order to determine whether a current period of absence is abnormally long or not.

If the person is present, the processing circuit can determine targeted information such as temporal information indicating a DET HLEV sunrise time (S33) and/or a DET HCOU bedtime (S32).

More precisely, the time of getting up (respectively the time of going to bed) are estimated as corresponding to the first (respectively the last) significant use of electricity in the home after (respectively before) a rest period. This makes it possible to follow the person's circadian rhythm.

During the rest period, the power consumption level is usually less than or equal to the passive consumption level. The rising time is estimated as being the first half hour with electrical activity, that is to say a level of electrical consumption higher than the level of passive consumption, from a predetermined time (for example 4 hours morning) or from bedtime the night before.

Bedtime is estimated as the last half hour with electrical activity from a predetermined time, for example 5 p.m.

Thus, each day, targeted information is determined from the recovered daily load curve and a daily history can thus be created.

If the person is absent, then the determination of targeted information such as a waking time, a bedtime, etc. is not applicable, so no calculation and recording of targeted information of this type is then carried out for a period of absence.

On the basis of the determined target information (presence indicator and temporal information of the type of a waking time and the type of a bedtime), the processing circuit monitors SURV (S40) the evolution over time of the alternation of periods of presence and absence of the person, as well as the person's circadian rhythm.

This evolution can be monitored in different ways, the simplest being to store the determined targeted information and allow it to be viewed, for example, by the person upon request.

Other ways to monitor this development are discussed below.

We now refer to the [ Fig. 3 ] which illustrates another example of carrying out a monitoring method.

The processing circuit obtains daily OBT SERIE (S10) a load curve as described above.

The processing circuit obtains OBT REP (S16) an indication of a usual meal time. Such an indication can for example come from a statistical study carried out for a given population. Such a study was, for example, carried out in France by INSEE.

Alternatively, such an indication may be a historical indication of a meal time of the person in the past.

The processing circuit searches for all moments of high power consumption during the past day. For this, the processing circuit can for example comparing COMP CONSO/REP (S22) the consumption values resulting from the daily load curve obtained with a threshold corresponding to a consumption level of cooking equipment.

From the comparisons made, the processing circuit can determine DET INFO REP (S34) temporal information of meal cooking time(s) during the past day. This determination can be refined on the basis of the indication of a usual meal time which makes it possible to select, among the moments of high electricity consumption, those most probably corresponding to the operation of cooking equipment.

Alternatively, the processing circuit can simply determine a meal intake indicator, that is to say estimate whether the person has had a meal or not during the day.

On the basis of the targeted information thus determined (temporal information of the type of a meal cooking hour or meal intake indicator), the processing circuit can monitor SURV (S40) the evolution over the time of a level of meal intake by the person. This development can be monitored in different ways, some of which are discussed further below.

We now refer to the [ Fig. 4 ] which illustrates another example of carrying out a monitoring method.

The processing circuit obtains daily OBT SERIE (S10) a load curve as described above.

The processing circuit obtains OBT ECS (S17), based on the load curve obtained, an estimate of the electricity consumption of a water heater of the electrical installation during the past day. It is in fact known, from the document FR3017975 A1 in particular, to estimate the electrical consumption of a given piece of equipment, here the water heater among a set of electrical equipment (i.e. the electrical installation) on the basis of such a load curve. The estimate is made by extracting the consumption of the water heater, called the “DHW part of the curve”, according to a process comprising several stages, including the main one is a difference operation between the load curve obtained and a lower “moving minimum” envelope of the load curve.

This estimate can be derived from measurements of the electrical consumption of such a water heater taken in isolation. Alternatively, this estimate can be determined by extracting from load curves of the electrical installation corresponding to previous days.

The electricity consumption of the water heater during the past day is a time series of values corresponding to the electricity consumption of the water heater during each of the consecutive time intervals forming the past day.

The processing circuit compares COMP CONSO/ECS (S23) the electrical consumption of the water heater, thus determined, with a threshold corresponding to passive electrical consumption of the water heater. The passive electricity consumption of the water heater can be taken from a database provided for example by the manufacturer of the water heater.

• d'obtenir une courbe de charge électrique de référence de l'installation électrique au cours d'une journée de référence correspondant à une période d'absence de la personne,
• d'extraire, à partir de la courbe de charge électrique de référence, la consommation électrique du chauffe-eau pour cette journée de référence, et
• de déterminer la consommation électrique passive du chauffe-eau comme étant égale à la consommation électrique du chauffe-eau ainsi extraite pour la journée de référence.

Alternatively, it can be planned:

• to obtain a reference electrical load curve of the electrical installation during a reference day corresponding to a period of absence of the person,
• to extract, from the reference electrical load curve, the electrical consumption of the water heater for this reference day, and
• to determine the passive electrical consumption of the water heater as being equal to the electrical consumption of the water heater thus extracted for the reference day.

Each elapsed day or elapsed reference day is a 24-hour period that may be subject to a peak/off-peak hours regime. For example, each load curve can cover time intervals of 30 minutes from day d at 10 p.m. to day d+1 at 10 p.m. in the case of an off-peak hours regime 10 p.m. - 4 a.m. / 12 p.m. - 2 p.m. . In the case of a peak/off-peak regime, the electricity consumption of the water heater is distributed as a priority, preferably during off-peak hours. Thus, it may be sufficient to compare only the water heater power consumption values for the time intervals corresponding to off-peak hours with the water heater passive power consumption values.

The processing circuit determines DET INFO ECS (S35) information indicative of domestic hot water consumption by the person during the past day.

The consumption of domestic hot water thus determined is recorded daily by the treatment circuit, which monitors SURV (S40) the evolution over time in order to evaluate the evolution of the person's level of hygiene over the course of the time.

We now refer to the [ Fig. 5 ] which illustrates another example of carrying out a monitoring method.

• obtient OBT SERIE (S10) une série temporelle indiquant, pour 48 intervalles de temps de 30 minutes consécutifs, une valeur de consommation électrique,
• compare COMP CONSO (S20) chacune de ces valeurs obtenues avec au moins une valeur de référence, et
• détermine DET INFO (S30) au moins une information ciblée.

Daily, the processing circuit:

• obtains OBT SERIE (S10) a time series indicating, for 48 consecutive 30-minute time intervals, a power consumption value,
• compares COMP CONSO (S20) each of these values obtained with at least one reference value, and
• determines DET INFO (S30) at least one targeted piece of information.

Obviously, as in all of the embodiments described, the number and duration of the time intervals are specified solely for educational purposes and are only indicative.

Furthermore, the nature of the reference values and the nature of the targeted information are irrelevant for the implementation of the embodiment illustrated in [ Fig. 5 ]. They can for example correspond to the reference values and the targeted information described in any of the embodiments above.

At each end of the week, the processing circuit has OBT GRP INFO (S36) a group of targeted information comprising, for each type of targeted information, up to one value per day of the past week. For each type of targeted information, a weekly average can for example be calculated.

The groups of targeted information previously obtained OBT GRP REF (S37) by the processing circuit at the end of the previous weeks form one or more groups of reference targeted information. The targeted reference information groups make it possible to know the person's lifestyle habits via their temporality of electricity consumption in their home.

The processing circuit monitors SURV (S40) the evolution over time of the targeted information included in the group of targeted information and the group(s) of reference targeted information.

To do this, for each type of targeted information, the latest weekly values are compared with usual historical reference values which can for example be obtained by statistical processing of the group of reference targeted information.

For example, if we consider as a type of targeted information temporal information indicating a waking time of the person, an example of a reference value can be an average or a median of the waking hours of the person during previous months.

A selection of the person's waking times during previous months can be determined by excluding certain extreme values exceeding a threshold in terms of variance. An example of a reference value may be an average or median of such a selection.

The processing circuit can then detect DET RUPT (S41) a break if the targeted information is stable before a very significant and immediate evolution (in one week) then becomes stable again at a different value.

The processing circuit can also detect DET EVOL (S42) a slow evolution which occurs when the targeted information, after having remained stable, begins to drift continuously for a long period of time (several weeks).

Generally speaking, the processing circuit can analyze the history of the weekly targeted information with a view to detecting with each new week added whether a significant event has occurred, such as a break or a significant slow evolution.

The processing circuit can then generate GEN ALERT (S50) an alert intended for a user who may be the person, a relative or an emergency service in the event of detection of such a significant event.

The alert can take different forms, such as an electronic message, a notification, a visual or audible alert, etc.

If no significant event is detected, it is also possible to indicate this to a user by generating information, such as a notification.

Furthermore, the usual historical references, being calculated from each history of targeted information, can be updated by the processing circuit each week so as to always integrate the last group of targeted information obtained.

Claims (15)

1. A method for monitoring a change over time of at least one activity of a person in a dwelling equipped with an electrical installation comprising at least one electrical equipment, the method being executed by a processing circuit and comprising:

   a) obtaining (S10), for a time period subdivided into a plurality of consecutive time intervals, a time series comprising, for each time interval, a value, associated with said time interval, indicating a level of consumption of at least one piece of electrical equipment of the electrical installation during said time interval,

   b) for each time interval, comparing (S20) the value associated with said time interval with at least one reference value,
   the method further comprising, for at least one activity of the person:

   c) based on the comparisons, determining (S30) targeted information indicative of said person's activity over the time period, and

   d) monitoring (S40) a change over time of said person's activity based on said targeted information and at least one targeted reference information item indicative of a usual activity of the person during a plurality of reference time periods preceding said time period,

   a method wherein at least one activity of the person is a presence or absence of the person during at least one portion of the time period, characterised in that:

   - the targeted information indicative of said presence or absence of the person is a targeted information item of the type of a presence indicator,

   - the at least one reference value includes a passive power consumption value,

   the method comprising:

   b1) for each time interval, comparing (S21) the value associated with said time interval with the passive power consumption value, and

   c1) for said presence or absence of the person, based on the comparisons, determining (S31) the targeted information of the type of a presence indicator so as to indicate the time period as being:

   - a period of absence of the person if each value associated with a time interval is less than or equal to the passive power consumption value, or

   - a period of presence of the person during the time period if at least one value associated with a time interval is greater than the passive power consumption value.
2. The monitoring method according to claim 1, further comprising:

   a1) obtaining (S11), for a historical time period subdivided into a plurality of consecutive historical time intervals, a historical time series comprising, for each historical time interval, a historical value, associated with said historical time interval, indicating a level of consumption of at least one piece of electrical equipment of the electrical installation during said historical time interval, and

   a3) estimating (S13), based on the historical time series, the passive power consumption value of said at least one piece of electrical equipment of the electrical installation.
3. The monitoring method according to claim 2, wherein estimating (S13), based on the historical time series, the passive power consumption value of said at least one electrical equipment of the electrical installation comprises:
   a2) implementing (S12) a statistical test to a factor between a first group formed of the historical values which are less than or equal to a threshold and a second group formed of the historical values which are greater than said threshold, the result of the statistical test being the estimated passive power consumption value of said at least one piece of electrical equipment of the electrical installation.
4. The monitoring method according to claim 3, wherein the statistical test is an ANOVA test.
5. The monitoring method according to one of claims 1 to 4, wherein:

   - the time period is indicated as being a period of presence of the person,

   - at least one activity of the person is a bedtime of the person,

   - the targeted information indicative of said bedtime of the person is a time information item of the type of a bedtime,

   the method comprising:
   c2) based on the comparisons, determining (S32) the time information of the type of a bedtime, said time information being associated with the last time interval of the time period from a predetermined hour such that the value associated with said time interval is greater than the passive power consumption value.
6. The monitoring method according to one of claims 1 to 5, wherein:

   - the time period is indicated as being a period of presence of the person,

   - at least one activity of the person is a wake up time of the person,

   - the targeted information indicative of said wake up time of the person is a time information item of the type of a wake up time,

   the method comprising:
   c3) based on the comparisons, determining (S33) the time information of the type of a wake up time, said time information being associated with the first time interval of the time period from a predetermined time such that the value associated with said time interval is greater than the passive power consumption value.
7. The monitoring method according to one of claims 1 to 6, wherein:

   - the electrical installation comprises at least one piece of electrical cooking equipment,

   - the activity of the person is an occurrence of meal intake by the person,

   - for each of the consecutive time intervals, the power consumption level during said time interval is a power consumption level of the electrical installation during said time interval,

   - the at least one reference power consumption level comprises a power consumption threshold of the electrical installation, indicative of an operation of at least one piece of electrical cooking equipment,

   the method comprising:

   b2) for each time interval, comparing the power consumption level of the electrical installation during said time interval with said consumption threshold,

   c4) based on the comparisons of the power consumption levels of the electrical installation with said consumption threshold, determining a targeted information item of the type of an indicator of meal intake by a person during at least one portion of the time period.
8. The monitoring method according to one of claims 1 to 7, wherein:

   - the electrical installation comprises an electric water heater,

   - the activity of the person is a consumption of domestic hot water from the electric water heater by the person,

   - for each of the consecutive time intervals, the power consumption level during said time interval is a power consumption level of the electric water heater during said time interval,

   - the at least one reference power consumption level includes an operating threshold of the electric water heater,

   the method comprising:

   b3) for each time interval, comparing the power consumption level of said electric water heater during said time interval with the operating threshold of said electric water heater,

   c5) based on the comparisons, determining a targeted information item of the type of an indicator of domestic hot water consumption from the electric water heater during the time period.
9. The monitoring method according to one of claims 1 to 8, comprising:

   a4) obtaining (S14) a group of targeted reference information comprising a plurality of targeted reference information of the same type as the determined targeted information, each targeted reference information of said type being indicative of an activity of the person during a past reference time period, prior to said time period,

   a5) obtaining (S15) a group of targeted information comprising, in addition to the determined targeted information, a plurality of previous targeted information of the same type as the determined targeted information, each previous targeted information being indicative of said activity of the person during a past time period, prior to said time period and subsequent to said past reference time periods, and

   wherein monitoring (S40) a change over time of an activity of the person comprises, for at least one determined targeted information, monitoring a change over time of said targeted information based on said group of targeted information and said group of targeted reference information.
10. The monitoring method according to claim 9, wherein monitoring a change over time of said targeted information comprises:
    d1) determining a difference between, on the one hand, the targeted information of said group of targeted information and, on the other hand, the targeted reference information of said group of targeted reference information, and, if said difference exceeds a predetermined difference threshold, detecting (S41) an event indicating a break in the activity of the person.
11. The monitoring method according to claim 9 or 10, wherein monitoring a change over time of said targeted information comprises:
    d2) determining a derivative of said targeted information with respect to time based on the group of targeted information and the group of targeted reference information, and, if said derivative exceeds a predetermined variation threshold, detecting (S42) an event indicating a change in the activity of the person.
12. The monitoring method according to claim 10 or 11, further comprising, after having monitored a change over time of said targeted information and detected an event,
    e) generating (S50) an alert in response to the detection of the event.
13. A computer program including instructions for implementing the method according to one of claims 1 to 12 when this program is executed by a processor.
14. A non-transitory computer-readable recording medium on which a program is recorded for implementing the method according to one of claims 1 to 12 when this program is executed by a processor.
15. A processing circuit comprising a processor 100 connected to a non-transitory recording medium 200 according to claim 14.

Images