Gas consumption prediction system and gas consumption prediction method

Abstract

To provide a gas consumption prediction device and a gas consumption prediction method capable of improving accuracy for predicting a future gas usage at a consumer's house.SOLUTION: A gas consumption prediction system comprises: a measurement unit that is installed in a customer's house and measuring a gas usage of each of a plurality of customer's houses; a collection unit that collects the gas usage used by each consumer's house from the measurement unit at a predetermined cycle; a classification unit that classifies each of the plurality of consumer's houses into one of a plurality of groups based on the gas usage status of the consumer's houses during a gas usage collection period; and a prediction unit that predicts a gas usage after the date of the last collection by the collection unit for each consumer's house based on a logic corresponding to each group classified by the classification unit.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a gas consumption prediction system and a gas consumption prediction method. [Background technology]

[0002] Conventionally, a meter reading data processing system has been disclosed that generates trend data representing the amount of change in gas usage or remaining gas over a certain period of time, and generates the timing for collecting the trend data based on the amount of change in gas usage or remaining gas (see Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 6263298 Summary of the Invention [Problem to be solved by the invention]

[0004] However, the meter reading data processing system of Patent Document 1 generates trend data for each consumer and predicts future gas consumption based on the past data of one consumer, which results in large variability in the predictions. Therefore, improving the accuracy of gas consumption predictions has been an issue.

[0005] Therefore, an object of the present invention is to provide a gas consumption prediction system and a gas consumption prediction method that can improve the accuracy of prediction of future gas consumption at a consumer's home compared to the conventional method. [Means for solving the problem]

[0006] The gas consumption prediction system of the present invention comprises a measurement unit installed in a consumer's home and measuring the gas usage of each of a plurality of consumer's homes; a collection unit that collects the gas usage of each consumer's home from the measurement unit at a predetermined interval; a classification unit that classifies each of the plurality of consumer's homes into one of a plurality of groups based on the gas usage at the consumer's home during the gas usage collection period; and a prediction unit that predicts the gas usage for each consumer's home from the date of the last collection by the collection unit based on logic corresponding to each of the groups classified by the classification unit.

[0007] According to the present invention, the classification unit classifies each of a plurality of consumer homes into one of a plurality of groups based on the gas usage status of the consumer homes during the collection period of the gas usage. This allows consumer homes with the same or similar gas usage trends to be classified into one group. That is, consumer homes with similar gas usage trends are classified into one group. Future gas usage is then predicted using logic corresponding to each group. In this way, unlike conventional systems that predict the gas usage of a consumer home based on the gas usage status of the consumer home, the prediction is made using an average logic that is optimal for each consumer home belonging to a group and that removes irregular gas usage, so that the prediction is not biased and the accuracy of the prediction is improved compared to conventional systems.

[0008] In the above invention, the classification unit may perform the classification based on the number of days without gas usage and the number of days with gas usage during the collection period as the gas usage status.

[0009] According to the above configuration, it becomes easier to classify consumer homes having the same or similar gas usage trends.

[0010] In the above invention, the prediction unit may have an artificial intelligence function and predict the gas usage amount using the artificial intelligence function with the gas usage amount for each consumer home collected by the collection unit as input.

[0011] According to the above configuration, by having the artificial intelligence continue machine learning or deep learning, more reliable predictions of gas usage can be made.

[0012] In the above invention, the prediction unit may predict the gas usage amount based on date-based information.

[0013] According to the above configuration, the gas consumption of the consumer's home is predicted based on date information such as the day of the week, holidays, seasons, etc., i.e., daily factors. This can further improve the reliability of the prediction of the gas consumption.

[0014] In the above invention, the prediction unit may predict the gas usage amount based on at least one of information based on a location of the customer's home and information based on a region in which the customer's home is located.

[0015] According to the above configuration, the gas usage of the customer's home is predicted based on information such as the address of the customer's home, legal measures (such as a state of emergency declaration) in the area where the customer's home is located, or weather in the area where the customer's home is located, that is, geographical factors. This makes it possible to further improve the reliability of the prediction of gas usage.

[0016] In the above invention, the prediction unit may predict the gas usage based on consumer home identification information which has at least one of information on gas equipment owned by the consumer home, family information of the consumer home, and occupation information of the consumer home.

[0017] According to the above configuration, the gas consumption of the consumer's home is predicted based on consumer's home specific information, such as information on the number of gas appliances owned by the consumer's home, information on the family structure of the consumer's home, or information on the consumer's occupation (e.g., restaurant, etc.) in the consumer's home, that is, consumer's home specific factors. This can further improve the reliability of the gas consumption prediction.

[0018] The gas consumption prediction method of the present invention is installed in a consumer's home, measures the gas usage of each of a plurality of consumer's homes, collects the gas usage of each of the consumer's homes at a predetermined interval, classifies each of the plurality of consumer's homes into one of a plurality of groups based on the gas usage by the consumer's home during the gas usage collection period, and predicts the gas usage for each consumer's home from the last collection date onwards based on the logic corresponding to each of the classified groups. Effect of the Invention

[0019] According to the present invention, it is possible to provide a gas consumption prediction system and a gas consumption prediction method that can improve the accuracy of prediction of future gas consumption in a consumer's home compared to the conventional art. [Brief description of the drawings]

[0020] [Figure 1] 1 is a block diagram showing a gas consumption prediction system according to an embodiment of the present invention. [Diagram 2] 2 is a diagram showing an example of a classification table stored in a storage unit in FIG. 1; [Diagram 3] 2 is a diagram showing an example of a first auxiliary information table stored in a storage unit in FIG. 1; FIG. [Figure 4] 1. FIG. 4 is a diagram showing an example of a second auxiliary information table stored in the storage unit of FIG. [Diagram 5] 1. FIG. 4 is a diagram showing an example of a third auxiliary information table stored in the storage unit of FIG. [Figure 6] 1 is a flowchart showing a logic generation flow. [Figure 7] 13 is a flowchart showing a prediction flow. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0021] Hereinafter, a gas consumption prediction system and a gas consumption prediction method according to an embodiment of the present invention will be described with reference to the drawings. The gas consumption prediction system and the gas consumption prediction method described below are merely one embodiment of the present invention. Therefore, the present invention is not limited to the following embodiment, and additions, deletions, and modifications are possible within the scope of the present invention.

[0022] Fig. 1 is a block diagram showing a gas consumption prediction system 100 according to one embodiment of the present invention. Fig. 2 is a diagram showing an example of a classification table Tb stored in the memory unit 4 of Fig. 1. As shown in Fig. 1, the gas consumption prediction system 100 includes a center server 10 and a plurality of consumer homes 20. Note that, although Fig. 1 illustrates four consumer homes 20, the gas consumption prediction system 100 may have a plurality of consumer homes 20, and may have three or less, or may have five or more.

[0023] Each consumer's home 20 is provided with two gas containers 11, a switch 12, a gas meter 13, and two gas appliances 14. The gas meter 13 corresponds to a measurement unit. Examples of consumer's homes 20 include, but are not limited to, hospitals, schools, municipal facilities, nursing homes, general homes, and commercial facilities, and the like. Consumer's homes 20 include any building in which gas can be used.

[0024] The gas container 11 is also called a gas cylinder, and is filled with a gas such as LP gas (liquefied petroleum gas). The switch 12 switches the gas supply path so that the gas in one of the two gas containers 11 is supplied to the consumer's house 20. As a result, even if one of the two gas containers 11 runs out of gas, the gas in the other gas container 11 can be supplied to the gas appliance 14.

[0025] The above-mentioned gas shortage is defined as follows. In this embodiment, when the gas container 11 to be used is switched by the switch 12, it means that the gas container 11 used before the switch is in a gas-out state. In this case, the switch 12 detects a change in pressure in the gas container 11 and switches the gas container 11. Specifically, the switch 12 applies pressure to the gas container 11 to fill it with gas, and when the gas runs out, the pressure drops. The switch 12 detects the change in pressure and switches the gas container 11 at the timing when the pressure drops. Therefore, in this embodiment, the drop in pressure detected by the switch 12 can be regarded as gas shortage. This definition is applied to each of the two gas containers 11 running out of gas.

[0026] The gas meter 13 measures the flow rate of gas supplied to each gas appliance 14. That is, the gas meter 13 measures the amount of gas used in the consumer's home 20. The gas meter 13 may take measures such as cutting off the gas supply path when it detects any abnormality. The gas meter 13 also includes a communication unit 13a. The communication unit 13a of the gas meter 13 has a function of wirelessly communicating with the center server 10. As a wireless communication method between the communication unit 13a and the center server 10, for example, a communication network such as the Internet, a LAN, or LPWA (Low Power Wide Area) can be used. The communication unit 13a periodically transmits information on the flow rate of gas used in the gas appliance 14 of the consumer's home 20 (information on gas usage) to the center server 10. The communication unit 13a may not be built into the gas meter 13, but may be configured as an external slave device and communicate with the gas meter 13.

[0027] The gas appliance 14 is, for example, a gas stove, a gas water heater, a gas fan heater, or the like, but is not limited to these, and the gas appliance 14 includes any appliance that consumes gas.

[0028] The center server 10 includes a collection unit 1, a classification unit 2, a prediction unit 3, a storage unit 4, and a communication interface (I / F) 5. Of the above components of the center server 10, the collection unit 1, the classification unit 2, and the prediction unit 3 are functionally realized by a microcontroller including a central processing unit (CPU) and memory (read only memory (ROM) and random access memory (RAM)) that stores a program, or an application specific integrated circuit (ASIC), etc. Also, the storage unit 4 can be various types of memory, a hard disk, etc.

[0029] The collection unit 1 receives and collects information on the flow rate of gas (gas usage amount) used by the gas appliances 14 of the consumer's home 20 at a predetermined cycle (e.g., one day cycle) from the communication unit 13a of the gas meter 13 in each consumer's home 20 via the communication interface 5. The collection unit 1 also stores the information on the gas usage amount in the memory unit 4. Note that, for example, an integrated value of the flow rate or a meter indicator value may be used as the information on the gas flow rate.

[0030] The storage unit 4 stores, in addition to information on the amount of gas used by the consumer's house 20, a first incidental information table Tb of Fig. 3, a second incidental information table Tc of Fig. 4, and a third incidental information table Td of Fig. 5. The first incidental information table Tb, the second incidental information table Tc, and the third incidental information table Td will be described in detail later.

[0031] The classification unit 2 classifies each of the multiple consumer homes 20 into one of multiple groups based on the gas usage status of the consumer homes 20 during a collection period of the gas usage amount. This will be described in detail below.

[0032] 2 is a diagram showing an example of the classification table Ta stored in the storage unit 4 of FIG. 1. The classification unit 2 reads out information on the amount of gas used for each consumer's home 20 stored in the storage unit 4, and performs classification based on, for example, the number of days without gas use and the number of days with gas use in a collection period as the gas usage status. In detail, the classification unit 2 calculates the ratio of the number of unused days in the collection period (i.e., the ratio of the number of days without gas use to the total number of days in the collection period) for each consumer's home 20 based on the number of days without gas use and the number of days with gas use in the collection period. In FIG. 2, three values ​​X, Y, and Z are shown as examples of the ratio of the number of unused days, and each of these X, Y, and Z may have a range.

[0033] The classification unit 2 determines which of X, Y, and Z in Fig. 2 the unused days ratio calculated as above corresponds to. Note that, although Fig. 2 shows three categories of X, Y, and Z as the unused days ratio divisions, this is not limiting, and the unused days ratio divisions may be two or four or more.

[0034] Next, the classification unit 2 extracts the maximum daily usage from the read gas usage information. The classification unit 2 then determines which of A, B, and C in FIG. 2 the extracted maximum daily usage corresponds to. In FIG. 2, three categories A, B, and C are shown as the classification of the daily gas usage, and each of these A, B, and C may have a range. Note that, although FIG. 2 shows three categories A, B, and C as the classification of the daily gas usage, the classification is not limited thereto, and the classification of the daily gas usage may be two or four or more.

[0035] Next, the classification unit 2 calculates the average daily usage from the read information on the gas usage. Then, the classification unit 2 determines which of L, M, and N in FIG. 2 the calculated average daily usage corresponds to. In FIG. 2, three categories L, M, and N are shown as categories of the average daily usage, and each of these L, M, and N may have a range. Note that, although FIG. 2 shows three categories L, M, and N as categories of the average daily usage, this is not limiting, and the average daily usage may be divided into two categories, or into four or more categories.

[0036] With the above configuration, the classification unit 2 classifies the consumer homes 20 into one group determined by a combination of the ratio of unused days, the maximum daily usage, and the average daily usage. Note that the consumer homes 20 to be classified as above may be extracted by removing those whose consumption patterns are obviously different, or by classifying them as general households or commercial businesses.

[0037] The prediction unit 3 predicts the amount of gas usage from the last collection date by the collection unit 1 for each consumer's home 20 based on logic (e.g., an arithmetic formula) corresponding to each group classified by the classification unit 2. In detail, the prediction unit 3 has an artificial intelligence 3a. The prediction unit 3 predicts the amount of gas usage from the last collection date using the artificial intelligence 3a, using data related to gas usage (the percentage of unused days obtained from gas usage information, the maximum daily usage, and the average daily usage) as input. The prediction by the artificial intelligence 3a of the prediction unit 3 is based on machine learning or deep learning.

[0038] The prediction unit 3 in this embodiment will be described in more detail. The logic is generated by the artificial intelligence 3a based on the classification conditions by the classification unit 2, and then the artificial intelligence 3a performs a prediction based on the logic. That is, the artificial intelligence 3a executes a logic generation process and a prediction process. In the logic generation process, the artificial intelligence 3a learns for each group using information for learning (information on gas usage collected from consumer homes 20 belonging to the group and additional information described below as necessary) as input, thereby generating the logic. In the prediction process, the artificial intelligence 3a predicts the gas usage after the last collection date for each of the specific consumer homes 20 by inputting information on gas usage and additional information collected from a specific consumer home 20 into the logic.

[0039] Here, in the above logic generation process, in addition to the gas usage amount, it is also possible to generate logic by adding additional information such as consumer home identification information including at least one of date-based information, location-based information, information on the gas appliances 14 owned by the consumer home 20, family information of the consumer home 20, and occupation information of the consumer home 20. In this way, by having the artificial intelligence 3a learn various additional information that is thought to affect gas consumption in addition to the gas usage amount information, the prediction accuracy can be further improved.

[0040] Next, a logic generation flow and a prediction flow in the gas consumption prediction system 100 of this embodiment will be described with reference to flowcharts. Fig. 6 is a flowchart showing the logic generation flow, and Fig. 7 is a flowchart showing the prediction flow.

[0041] 6, the gas meter 13 measures the amount of gas used in the consumer's home 20 (step S1). Next, the collection unit 1 receives and collects information on the amount of gas used in the consumer's home 20 at a predetermined interval from the communication unit 13a of the gas meter 13 in each consumer's home 20 (step S2). In this case, for example, an integrated value of the flow rate or a meter pointer value may be used as the information on the amount of gas used.

[0042] Next, the collection unit 1 collects the additional information (step S3). Next, the classification unit 2 classifies each of the multiple consumer homes 20 into one of multiple groups based on the gas usage by the consumer homes 20 during the collection period of the gas usage and the additional information (step S4).

[0043] Next, the artificial intelligence 3a of the prediction unit 3 generates logic corresponding to each group classified by the classification unit 2 (step S5).

[0044] 7, the gas meter 13 measures the amount of gas used in the consumer's home 20 (step S11). Next, the collection unit 1 collects information on the amount of gas used in the consumer's home 20 (step S12) and collects additional information (step S13).

[0045] Next, the classification unit 2 classifies each of the multiple consumer homes 20 into one of multiple groups based on the gas usage status and additional information of the consumer homes 20 during the collection period of the gas usage amount (step S14).

[0046] Then, the artificial intelligence 3a of the prediction unit 3 predicts the amount of gas usage from the date of the last collection by the collection unit 1 for each consumer's home 20 based on the logic corresponding to each group classified by the classification unit 2 (step S15).

[0047] The logic generation process and the forecast generation process must use the same type of data. In other words, the data formats (e.g., gas consumption, date, temperature, data division method) used in the logic generation flow and the forecast flow are the same, but the data itself is different.

[0048] In addition, the prediction flow uses past data from the base point to be predicted (for example, one year from the base point), while the logic generation flow uses data from a period different from that of the prediction flow (data from the past year that precedes the data used in the prediction flow).

[0049] Furthermore, the logic generation process is updated periodically, such as once a year, and also when appropriate, such as when there is a major change in gas usage due to a large-scale disaster or a change in lifestyle.

[0050] The effect of the above-mentioned incidental information on the gas usage may be provided as a coefficient in a table or the like to correct the predicted gas usage, and although the accuracy is lower than that of the above-mentioned learning, a similar effect can be obtained. Below, a specific example is given to explain the method of performing correction based on incidental information using a coefficient.

[0051] As in the following specific example, the prediction unit 3 may add additional information when predicting the gas usage amount for each consumer's house 20. Fig. 3 is a diagram showing an example of a first additional information table Tb stored in the storage unit 4 of Fig. 1, Fig. 4 is a diagram showing an example of a second additional information table Tc stored in the storage unit 4 of Fig. 1, and Fig. 5 is a diagram showing an example of a third additional information table Td stored in the storage unit 4 of Fig. 1.

[0052] 3, in the first additional information table Tb, a holiday coefficient is defined for each type of consumer residence 20. Taking into consideration that the amount of gas usage on weekdays and holidays (Saturdays, Sundays, and public holidays) differs depending on the type of consumer residence 20, the holiday coefficient is defined for each consumer residence 20 as the relative amount of gas usage on holidays based on the average amount of gas usage on weekdays.

[0053] For example, as shown in FIG. 3, in a hospital, gas usage on holidays is expected to be less than that on weekdays, so the holiday coefficient for the hospital can be set to, for example, 0.9. On the other hand, in a commercial facility, gas usage on holidays is expected to be greater than that on weekdays, so the holiday coefficient for the commercial facility can be set to, for example, 2.2. The holiday coefficients 0.9 and 2.2 are merely examples and can be set arbitrarily. In such a configuration, the prediction unit 3 may multiply the gas usage related to holidays among the gas usage predicted by the above logic (arithmetic formula) by the holiday coefficient in the first incidental information table Tb. This makes it possible to obtain a prediction value with improved reliability. Note that instead of or in addition to the holiday coefficient, gas usage may be predicted using a coefficient that takes into account, for example, the season.

[0054] Next, as shown in FIG. 4, in the second additional information table Tc, an address coefficient, a region coefficient, and a weather coefficient are determined for each consumer home 20. In the second additional information table Tc, k1 to k8 are predetermined coefficients. The address coefficient is determined for each consumer home 20, taking into consideration that the gas consumption varies due to, for example, differences in average temperature for each address. In this case, if the address of the consumer home 20 is, for example, in the Kyushu region, the gas consumption in the gas appliances 14, such as a gas water heater or a gas fan heater, is expected to be relatively low, so the address coefficient can be set to a relatively small value. On the other hand, if the address of the consumer home 20 is, for example, in Hokkaido, the gas consumption in the gas appliances 14 is expected to be relatively high, so the address coefficient can be set to a relatively large value.

[0055] The regional coefficient is determined for each consumer's residence 20, taking into consideration differences in gas usage due to legal measures (such as a state of emergency declaration) in the region, for example. In this case, if a legal measure such as a state of emergency declaration has been issued in the region of the consumer's residence 20, the consumer is expected to spend a relatively long time at home in the consumer's residence 20, so the regional coefficient can be set to a relatively large value. On the other hand, if a legal measure such as a state of emergency declaration has not been issued in the region of the consumer's residence 20, the consumer is expected to spend a relatively short time at home in the consumer's residence 20, so the regional coefficient can be set to a relatively small value.

[0056] The weather coefficient is a coefficient similar to the address coefficient, but is determined for each consumer's home 20, taking into consideration that gas consumption differs between areas with a lot of rain and snow and areas with little rain and snow. In this case, if the consumer's home 20 is located in an area with a lot of rain and snow, the gas consumption in the gas appliance 14 is expected to be relatively high, and the weather coefficient can be set to a relatively large value. On the other hand, if the consumer's home 20 is located in an area with little rain and snow, the gas consumption in the gas appliance 14 is expected to be relatively low, and the weather coefficient can be set to a relatively small value. As shown in FIG. 4, a maximum value k5 and a minimum value k6 may be determined for the weather coefficient of consumer's home A, and a maximum value k7 and a minimum value k8 may be determined for the weather coefficient of consumer's home B, and the maximum and minimum values ​​may be used according to the significant climate change in recent years.

[0057] In the above configuration, the prediction unit 3 may multiply the gas usage predicted by the above logic (arithmetic formula) by at least one of an address coefficient, a region coefficient, and a weather coefficient. This makes it possible to obtain a prediction value with improved reliability.

[0058] Alternatively, the following additional information may be adopted. As shown in FIG. 5, in the third additional information table Td, a number coefficient, a family coefficient, and an occupation coefficient are determined for each consumer home 20. In the third additional information table Td, k11 to k16 are predetermined coefficients. The number coefficient is determined for each consumer home 20, taking into consideration that the consumer home 20 that consumes more gas uses a gas meter 13 with a number corresponding to a larger capacity of the gas container 11. In this case, if the number of the gas meter 13 is relatively large, the gas consumption in the gas appliance 14 is expected to be relatively large, so the number coefficient can be set to a relatively large value. On the other hand, if the number of the gas meter 13 is relatively small, the gas consumption in the gas appliance 14 is expected to be relatively small, so the number coefficient can be set to a relatively small value.

[0059] The family coefficient is determined for each consumer home 20, taking into consideration that the amount of gas usage differs depending on the number of family members in the consumer home 20. In this case, if the number of family members in the consumer home 20 is relatively large, the amount of gas consumption in the gas appliances 14 is expected to be relatively large, so the family coefficient can be set to a relatively large value. On the other hand, if the number of family members in the consumer home 20 is relatively small, the amount of gas consumption in the gas appliances 14 is expected to be relatively small, so the family coefficient can be set to a relatively small value.

[0060] The occupation coefficient is determined for each consumer home 20, taking into consideration that the amount of gas consumption differs depending on the occupation of the consumer at the consumer home 20. In this case, if the consumer home 20 is, for example, a restaurant, the amount of gas consumption in the gas appliances 14 is expected to be relatively large, so the occupation coefficient can be set to a relatively large value. On the other hand, if the consumer home 20 is, for example, an ordinary household, the amount of gas consumption in the gas appliances 14 is expected to be relatively small, so the occupation coefficient can be set to a relatively small value.

[0061] In the above configuration, the prediction unit 3 may multiply the gas usage predicted by the above logic (arithmetic formula) by at least one of the number coefficient, the family coefficient, and the occupation coefficient. This makes it possible to obtain a prediction value with improved reliability.

[0062] As described above, according to this embodiment, the classification unit 2 classifies each of the multiple consumer homes 20 into one of multiple groups based on the gas usage status of the consumer homes 20 during the collection period of the gas usage. This allows consumer homes 20 with the same or similar gas usage trends to be classified into one group. That is, consumer homes 20 with similar gas usage trends are classified into one group. Then, future gas usage is predicted using logic corresponding to each group. In this way, unlike the conventional system that predicts the gas usage of one consumer home 20 based on the gas usage status of the consumer home 20, the prediction is made using an average logic that is optimal for each consumer home 20 belonging to one group and that excludes irregular gas usage, so that the prediction is not biased and the accuracy of the prediction is improved compared to the conventional system.

[0063] (Modification) The present invention is not limited to the above-described embodiment, and various modifications are possible without departing from the gist of the present invention. For example, the following modifications are possible.

[0064] In the above embodiment, the artificial intelligence 3a of the prediction unit 3 is configured to predict the amount of gas usage after the date of the last collection by the collection unit 1 based on logic (arithmetic formula) corresponding to each group classified by the classification unit 2, but this is not limited to this. The prediction unit 3 may predict the amount of gas usage using, for example, a table or a function graph showing the relationship between the data collected by the collection unit 1 and the predicted value corresponding to the collected data, as the logic.

[0065] In the above embodiment, the gas meter 13 transmits information on the flow rate of gas used by the gas appliance 14 in the consumer's house 20 to the center server 10 via wireless communication, but the present invention is not limited to this. The gas meter 13 may transmit information on the flow rate of gas to the center server 10 via wired communication.

[0066] In addition, in the above embodiment, two gas containers 11 and two gas appliances 14 are exemplified, but this is not limited to this, and the number of gas containers 11 may be one or three or more, and the number of gas appliances 14 may be one or three or more.

[0067] Furthermore, in the above embodiment, the gas container 11 is filled with LP gas (liquefied petroleum gas) as an example of a gas, but this is not limited thereto, and the gas container 11 may be filled with other gases such as oxygen. [Explanation of symbols]

[0068] 1 Collection Department 2 Classification section 3. Prediction Department 3a Artificial intelligence 4 Storage section 10. Central Server 11 Gas bottles 13 Gas meter 14 Gas appliances 20 Customer house 100 Gas Consumption Prediction System

Claims

1. A measuring unit that is installed in a consumer home and measures the amount of gas used by each of the consumer homes; a collection unit that collects the gas usage amount of each customer house from the measurement unit at a predetermined period; a classification unit that classifies each of the plurality of consumer homes into one of a plurality of groups based on a gas usage status of the consumer home during a collection period of the gas usage amount; and a prediction unit that predicts the amount of gas usage for each consumer home after the date of the last collection by the collection unit based on consumer home identification information having at least one of information on gas appliances owned by the consumer home classified by the classification unit, family information on the consumer home, and occupation information on the consumer home, A gas consumption prediction system, wherein the classification unit performs the classification based on the number of days without gas usage and the number of days with gas usage during the collection period as the gas usage status.

2. The gas consumption prediction system of claim 1, wherein the prediction unit has an artificial intelligence function and predicts the gas usage amount using the artificial intelligence function based on the gas usage amount for each consumer home collected by the collection unit as input.

3. The gas consumption prediction system according to claim 1 , wherein the prediction unit predicts the gas usage amount based on date-based information.

4. A gas consumption prediction system as described in any one of claims 1 to 3, wherein the prediction unit predicts the gas usage based on at least one of information based on the location of the consumer's home and information based on the area in which the consumer's home is located.

5. A gas consumption prediction system as described in any one of claims 1 to 4, wherein the classification unit performs the classification based on at least one of the ratio of the number of days without gas usage to the total number of days in the collection period, the maximum daily usage, and the average daily usage.

6. A gas consumption prediction system as described in any one of claims 1 to 5, wherein the classification unit classifies each of the multiple consumer homes into one of multiple groups based on consumer home identification information having at least one of information on gas equipment owned by the consumer home, family information of the consumer home, and occupational information of the consumer home.

7. It is installed in the customer's home and measures the gas usage of each customer's home. Collecting gas usage amounts for each of the consumer homes at a predetermined interval; classifying each of the plurality of consumer homes into one of a plurality of groups based on the number of days without gas usage and the number of days with gas usage during the collection period as a gas usage status by the consumer home during the collection period of the gas usage amount; A gas consumption prediction method that predicts the amount of gas usage for each consumer home from the last collection date based on consumer home identification information that has at least one of information on gas equipment owned by the classified consumer home, family information of the consumer home, and occupational information of the consumer home.

Images